IT_Henry_sunny
IT_Henry_sunny

使用TensorFlow keras搭建神经网络实现牛奶质量分类

本文通过TensorFlow框架中的keras构建简单神经网络对牛奶质量进行分类（预测），具有较高的准确率，下面进行详细的介绍。

文章目录

    • 数据集介绍
    • 分类过程介绍
    • 总结与分析

数据集介绍

数据集的来源为kaggle网站，地址为https://www.kaggle.com/datasets/cpluzshrijayan/milkquality
在此感谢数据集的作者，大家可自行下载。

  • 该数据集为人工观测收集，可用于构建机器学习模型来预测牛奶的质量。
  • 该数据集由pH值、温度、味道、气味、脂肪、浊度和颜色7个特征组成。
  • 在一般情况下，牛奶质量取决于这些特征参数，可用于预测牛奶的质量。

标签值分为3类，分别是low、medium、high，以此来区分牛奶的质量。

分类过程介绍

首先读取实验数据集并进行处理
由于数据集标签列中为字符串类型，为了便于实验，首先将标签值转换为数值类型。
转换的规则为：
low = 0
medium = 1
high = 2
处理后输出数据集

import pandas as pd
import tensorflow as tf
import numpy as np

# 读取数据集并处理
read_data = pd.read_csv('milknew.csv')
# 读取进来的csv文件
print('显示读取进来的csv文件：\n', read_data)
# 使用to_string()可以输出全部数据，否则只能输出前五行和后五行数据，中间数据以...代替
# print(read_data.to_string())
# 读取csv文件值
print('读取csv文件的值：\n', read_data.values)
# 读取csv文件中'Grade'列的值
data_Grade = read_data['Grade'].values
print('读取csv文件中【Grade】列的值：\n', data_Grade)
# 将csv文件中'Grade'列的值由字符串型变为数值型
'''
low = 0
medium = 1
high = 2
'''
for _, grade in enumerate(data_Grade):
    # print(_, grade)
    if grade == 'low':
        data_Grade[_] = 0
    elif grade == 'medium':
        data_Grade[_] = 1
    elif grade == 'high':
        data_Grade[_] = 2
print('更新后csv文件中【Grade】列的值：\n',data_Grade)
read_data['Grade'] = data_Grade
#print(read_data['Grade'])
# 处理后的数据集文件
datasets = read_data
print('处理后的数据集文件：\n', datasets)
# print(f'数据集有{datasets.shape[0]}行')
# print(f'数据集有{datasets.shape[1]}列')

显示读取进来的csv文件：
        pH  Temprature  Taste  Odor  Fat   Turbidity  Colour   Grade
0     6.6          35      1     0     1          0     254    high
1     6.6          36      0     1     0          1     253    high
2     8.5          70      1     1     1          1     246     low
3     9.5          34      1     1     0          1     255     low
4     6.6          37      0     0     0          0     255  medium
...   ...         ...    ...   ...   ...        ...     ...     ...
1054  6.7          45      1     1     0          0     247  medium
1055  6.7          38      1     0     1          0     255    high
1056  3.0          40      1     1     1          1     255     low
1057  6.8          43      1     0     1          0     250    high
1058  8.6          55      0     1     1          1     255     low

[1059 rows x 8 columns]
读取csv文件的值：
 [[6.6 35 1 ... 0 254 'high']
 [6.6 36 0 ... 1 253 'high']
 [8.5 70 1 ... 1 246 'low']
 ...
 [3.0 40 1 ... 1 255 'low']
 [6.8 43 1 ... 0 250 'high']
 [8.6 55 0 ... 1 255 'low']]
读取csv文件中【Grade】列的值：
 ['high' 'high' 'low' ... 'low' 'high' 'low']
更新后csv文件中【Grade】列的值：
 [2 2 0 ... 0 2 0]
处理后的数据集文件：
        pH  Temprature  Taste  Odor  Fat   Turbidity  Colour Grade
0     6.6          35      1     0     1          0     254     2
1     6.6          36      0     1     0          1     253     2
2     8.5          70      1     1     1          1     246     0
3     9.5          34      1     1     0          1     255     0
4     6.6          37      0     0     0          0     255     1
...   ...         ...    ...   ...   ...        ...     ...   ...
1054  6.7          45      1     1     0          0     247     1
1055  6.7          38      1     0     1          0     255     2
1056  3.0          40      1     1     1          1     255     0
1057  6.8          43      1     0     1          0     250     2
1058  8.6          55      0     1     1          1     255     0

[1059 rows x 8 columns]

便于实验方便，将上述数据集中的特征列与标签列分别提取出来，在提取过程中发现标签列为object类型，不便于后续训练模型，所以此处将标签列转换为int类型。

# 将训练集的特征和标签提取出来
# 提取特征列
features_data = datasets.iloc[:,:-1].values # 提取全部行，除最后一列的数据
# 提取标签列
labels_data = datasets.iloc[:,-1:].values # 提取全部行，最后一列的数据
# 下面这种语句也可以实现提取相应的列
# features_data = datasets.iloc[:,:7].values # 提取全部行，前7列的数据
# labels_data = datasets.iloc[:,7:8].values # 提取全部行，最后一列的数据

# print('数据集特征列：\n', features_data)
print('数据集特征列的类型：\n', features_data.dtype)
# print('数据集标签列：\n', labels_data)
print('数据集标签列的类型：\n', labels_data.dtype)
# labels_data.dtype的类型为object，后续训练需要转化为int类型

# 建立一个空的numpy一维矩阵，以便于将labels_data从object型转成int型，用于后续模型训练
labels_data_numpy = np.zeros((1, len(labels_data)))

# 将labels_data数字装进numpy一维矩阵中
for _, label in enumerate(labels_data):
    # print(_, label[0])
    labels_data_numpy[0][_] = label[0]

print("类型转化后的标签列：\n", labels_data_numpy[0])
print("类型转化后的标签列的类型：\n", labels_data_numpy[0].dtype)

# labels_data_np承接类型转化后的标签列，以便于后续操作
labels_data_np = labels_data_numpy[0]

# 将float64类型转化为int32类型
labels_data_np = tf.cast(labels_data_np, tf.int32)

# 为了方便后续操作，依然将处理好的标签列命名为labels_data
labels_data = labels_data_np

# 输出数据集处理后的结果
print('数据集特征列：\n', features_data)
print('数据集特征列的类型为：\n', features_data.dtype)
print('数据集标签列：\n', labels_data)
print('数据集标签列的类型为：\n', labels_data.dtype)

数据集特征列的类型：
 float64
数据集标签列的类型：
 object
类型转化后的标签列：
 [2. 2. 0. ... 0. 2. 0.]
类型转化后的标签列的类型：
 float64
数据集特征列：
 [[  6.6  35.    1.  ...   1.    0.  254. ]
 [  6.6  36.    0.  ...   0.    1.  253. ]
 [  8.5  70.    1.  ...   1.    1.  246. ]
 ...
 [  3.   40.    1.  ...   1.    1.  255. ]
 [  6.8  43.    1.  ...   1.    0.  250. ]
 [  8.6  55.    0.  ...   1.    1.  255. ]]
数据集特征列的类型为：
 float64
数据集标签列：
 tf.Tensor([2 2 0 ... 0 2 0], shape=(1059,), dtype=int32)
数据集标签列的类型为：

下面进行网络结构的搭建，配置和训练模型
网络结构有3层全连接层：
第一层有30个神经元，激活函数为relu
第二层有90个神经元，激活函数为relu
第三层，也是输出层，有3个神经元，激活函数为softmax
网络配置中：
使用Adagrad优化器
学习率为0.1
使用稀疏多分类交叉熵损失函数(SparseCategoricalCrossentropy)
训练过程中：
batch_size设置为32一组
训练轮数epochs设置为500轮
每轮迭代前随机打乱数据样本
验证集占数据集的比例为0.2
每100轮使用验证集验证一次

# 搭建网络结构
model = tf.keras.models.Sequential([
    # 这是一个全连接层，3是神经元个数，kernel_regularizer是选用的正则化方法 
    tf.keras.layers.Dense(30, activation='relu', kernel_regularizer=tf.keras.regularizers.l2()),  
    tf.keras.layers.Dense(90, activation='relu', kernel_regularizer=tf.keras.regularizers.l2()), 
    tf.keras.layers.Dense(3, activation='softmax', kernel_regularizer=tf.keras.regularizers.l2()) 
])
print('网络结构搭建：\n', model)

# 配置训练方法
model.compile(optimizer=tf.keras.optimizers.Adagrad(learning_rate=0.1),
             loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=False),
             metrics=['sparse_categorical_accuracy'])

# 执行训练过程
# validation_split是测试集的比例
# validation_freq表示每迭代多少次训练集就在验证集中验证一次准确率
# batch_size默认为32，也可改成别的数字，一般设置成2的n次方
model.fit(features_data, labels_data ,batch_size=32,epochs=500,shuffle=True,validation_split=0.2,validation_freq=100)

网络结构搭建：
 
Epoch 1/500
27/27 [==============================] - 0s 1ms/step - loss: 22.1251 - sparse_categorical_accuracy: 0.4321
Epoch 2/500
27/27 [==============================] - 0s 997us/step - loss: 1.6387 - sparse_categorical_accuracy: 0.4723
Epoch 3/500
27/27 [==============================] - 0s 1ms/step - loss: 1.3700 - sparse_categorical_accuracy: 0.5301
Epoch 4/500
27/27 [==============================] - 0s 1ms/step - loss: 1.3302 - sparse_categorical_accuracy: 0.5289
Epoch 5/500
27/27 [==============================] - 0s 1ms/step - loss: 1.2542 - sparse_categorical_accuracy: 0.5207
Epoch 6/500
27/27 [==============================] - 0s 1ms/step - loss: 1.2033 - sparse_categorical_accuracy: 0.5230
Epoch 7/500
27/27 [==============================] - 0s 1ms/step - loss: 1.1840 - sparse_categorical_accuracy: 0.5336
Epoch 8/500
27/27 [==============================] - 0s 1ms/step - loss: 1.1732 - sparse_categorical_accuracy: 0.5348
Epoch 9/500
27/27 [==============================] - 0s 1ms/step - loss: 1.1447 - sparse_categorical_accuracy: 0.5348
Epoch 10/500
27/27 [==============================] - 0s 1ms/step - loss: 1.1235 - sparse_categorical_accuracy: 0.5537
Epoch 11/500
27/27 [==============================] - 0s 1ms/step - loss: 1.0977 - sparse_categorical_accuracy: 0.5490
Epoch 12/500
27/27 [==============================] - 0s 892us/step - loss: 1.0848 - sparse_categorical_accuracy: 0.5632
Epoch 13/500
27/27 [==============================] - 0s 1ms/step - loss: 1.0857 - sparse_categorical_accuracy: 0.5490
Epoch 14/500
27/27 [==============================] - 0s 1ms/step - loss: 1.0787 - sparse_categorical_accuracy: 0.5691
Epoch 15/500
27/27 [==============================] - 0s 1ms/step - loss: 1.0580 - sparse_categorical_accuracy: 0.5632
Epoch 16/500
27/27 [==============================] - 0s 1ms/step - loss: 1.0658 - sparse_categorical_accuracy: 0.5809
Epoch 17/500
27/27 [==============================] - 0s 1ms/step - loss: 1.0720 - sparse_categorical_accuracy: 0.5702
Epoch 18/500
27/27 [==============================] - 0s 1ms/step - loss: 1.0424 - sparse_categorical_accuracy: 0.5998
Epoch 19/500
27/27 [==============================] - 0s 1ms/step - loss: 1.0271 - sparse_categorical_accuracy: 0.5880
Epoch 20/500
27/27 [==============================] - 0s 1ms/step - loss: 1.0194 - sparse_categorical_accuracy: 0.6104
Epoch 21/500
27/27 [==============================] - 0s 1ms/step - loss: 1.0232 - sparse_categorical_accuracy: 0.6045
Epoch 22/500
27/27 [==============================] - 0s 1ms/step - loss: 1.0112 - sparse_categorical_accuracy: 0.6116
Epoch 23/500
27/27 [==============================] - 0s 958us/step - loss: 1.0126 - sparse_categorical_accuracy: 0.6234
Epoch 24/500
27/27 [==============================] - 0s 1ms/step - loss: 0.9846 - sparse_categorical_accuracy: 0.6257
Epoch 25/500
27/27 [==============================] - 0s 1ms/step - loss: 0.9905 - sparse_categorical_accuracy: 0.6139
Epoch 26/500
27/27 [==============================] - 0s 1ms/step - loss: 1.0140 - sparse_categorical_accuracy: 0.6057
Epoch 27/500
27/27 [==============================] - 0s 1ms/step - loss: 0.9873 - sparse_categorical_accuracy: 0.6269
Epoch 28/500
27/27 [==============================] - 0s 952us/step - loss: 0.9959 - sparse_categorical_accuracy: 0.6269
Epoch 29/500
27/27 [==============================] - 0s 1ms/step - loss: 0.9571 - sparse_categorical_accuracy: 0.6600
Epoch 30/500
27/27 [==============================] - 0s 1ms/step - loss: 0.9845 - sparse_categorical_accuracy: 0.6246
Epoch 31/500
27/27 [==============================] - 0s 1ms/step - loss: 0.9532 - sparse_categorical_accuracy: 0.6517
Epoch 32/500
27/27 [==============================] - 0s 1ms/step - loss: 0.9519 - sparse_categorical_accuracy: 0.6411
Epoch 33/500
27/27 [==============================] - 0s 1ms/step - loss: 0.9446 - sparse_categorical_accuracy: 0.6730
Epoch 34/500
27/27 [==============================] - 0s 1ms/step - loss: 0.9493 - sparse_categorical_accuracy: 0.6458
Epoch 35/500
27/27 [==============================] - 0s 907us/step - loss: 0.9605 - sparse_categorical_accuracy: 0.6375
Epoch 36/500
27/27 [==============================] - 0s 1ms/step - loss: 0.9292 - sparse_categorical_accuracy: 0.6694
Epoch 37/500
27/27 [==============================] - 0s 1ms/step - loss: 0.9351 - sparse_categorical_accuracy: 0.6553
Epoch 38/500
27/27 [==============================] - 0s 1ms/step - loss: 0.9844 - sparse_categorical_accuracy: 0.6281
Epoch 39/500
27/27 [==============================] - 0s 1ms/step - loss: 0.9111 - sparse_categorical_accuracy: 0.6930
Epoch 40/500
27/27 [==============================] - 0s 1ms/step - loss: 0.9483 - sparse_categorical_accuracy: 0.6600
Epoch 41/500
27/27 [==============================] - 0s 1ms/step - loss: 0.9050 - sparse_categorical_accuracy: 0.6753
Epoch 42/500
27/27 [==============================] - 0s 1ms/step - loss: 0.9191 - sparse_categorical_accuracy: 0.6765
Epoch 43/500
27/27 [==============================] - 0s 1ms/step - loss: 0.9438 - sparse_categorical_accuracy: 0.6588
Epoch 44/500
27/27 [==============================] - 0s 956us/step - loss: 0.8847 - sparse_categorical_accuracy: 0.7273
Epoch 45/500
27/27 [==============================] - 0s 1ms/step - loss: 0.9330 - sparse_categorical_accuracy: 0.6635
Epoch 46/500
27/27 [==============================] - 0s 1ms/step - loss: 0.9123 - sparse_categorical_accuracy: 0.6824
Epoch 47/500
27/27 [==============================] - 0s 934us/step - loss: 0.8701 - sparse_categorical_accuracy: 0.7285
Epoch 48/500
27/27 [==============================] - 0s 1ms/step - loss: 0.8512 - sparse_categorical_accuracy: 0.7214
Epoch 49/500
27/27 [==============================] - 0s 905us/step - loss: 0.8901 - sparse_categorical_accuracy: 0.7226
Epoch 50/500
27/27 [==============================] - 0s 918us/step - loss: 0.9285 - sparse_categorical_accuracy: 0.6812
Epoch 51/500
27/27 [==============================] - 0s 1ms/step - loss: 0.8719 - sparse_categorical_accuracy: 0.7131
Epoch 52/500
27/27 [==============================] - 0s 1ms/step - loss: 0.8747 - sparse_categorical_accuracy: 0.7344
Epoch 53/500
27/27 [==============================] - 0s 1ms/step - loss: 0.8510 - sparse_categorical_accuracy: 0.7367
Epoch 54/500
27/27 [==============================] - 0s 1ms/step - loss: 0.8178 - sparse_categorical_accuracy: 0.7591
Epoch 55/500
27/27 [==============================] - 0s 1ms/step - loss: 0.8557 - sparse_categorical_accuracy: 0.7237
Epoch 56/500
27/27 [==============================] - 0s 962us/step - loss: 0.8976 - sparse_categorical_accuracy: 0.6930
Epoch 57/500
27/27 [==============================] - 0s 1ms/step - loss: 0.9161 - sparse_categorical_accuracy: 0.6812
Epoch 58/500
27/27 [==============================] - 0s 1ms/step - loss: 0.8566 - sparse_categorical_accuracy: 0.7285
Epoch 59/500
27/27 [==============================] - 0s 1ms/step - loss: 0.8546 - sparse_categorical_accuracy: 0.7414
Epoch 60/500
27/27 [==============================] - 0s 980us/step - loss: 0.8403 - sparse_categorical_accuracy: 0.7438
Epoch 61/500
27/27 [==============================] - 0s 997us/step - loss: 0.8893 - sparse_categorical_accuracy: 0.6777
Epoch 62/500
27/27 [==============================] - 0s 1ms/step - loss: 0.8692 - sparse_categorical_accuracy: 0.7096
Epoch 63/500
27/27 [==============================] - 0s 1ms/step - loss: 0.8341 - sparse_categorical_accuracy: 0.7296
Epoch 64/500
27/27 [==============================] - 0s 1ms/step - loss: 0.8447 - sparse_categorical_accuracy: 0.7332
Epoch 65/500
27/27 [==============================] - 0s 976us/step - loss: 0.8475 - sparse_categorical_accuracy: 0.7084
Epoch 66/500
27/27 [==============================] - 0s 971us/step - loss: 0.8363 - sparse_categorical_accuracy: 0.7414
Epoch 67/500
27/27 [==============================] - 0s 1ms/step - loss: 0.8759 - sparse_categorical_accuracy: 0.6848
Epoch 68/500
27/27 [==============================] - 0s 1ms/step - loss: 0.8390 - sparse_categorical_accuracy: 0.7296
Epoch 69/500
27/27 [==============================] - 0s 973us/step - loss: 0.7926 - sparse_categorical_accuracy: 0.7414
Epoch 70/500
27/27 [==============================] - 0s 989us/step - loss: 0.8728 - sparse_categorical_accuracy: 0.7013
Epoch 71/500
27/27 [==============================] - 0s 943us/step - loss: 0.8297 - sparse_categorical_accuracy: 0.7214
Epoch 72/500
27/27 [==============================] - 0s 961us/step - loss: 0.8263 - sparse_categorical_accuracy: 0.7285
Epoch 73/500
27/27 [==============================] - 0s 874us/step - loss: 0.8179 - sparse_categorical_accuracy: 0.7332
Epoch 74/500
27/27 [==============================] - 0s 908us/step - loss: 0.8657 - sparse_categorical_accuracy: 0.7119
Epoch 75/500
27/27 [==============================] - 0s 884us/step - loss: 0.7939 - sparse_categorical_accuracy: 0.7414
Epoch 76/500
27/27 [==============================] - 0s 914us/step - loss: 0.7630 - sparse_categorical_accuracy: 0.7733
Epoch 77/500
27/27 [==============================] - 0s 984us/step - loss: 0.7445 - sparse_categorical_accuracy: 0.7922
Epoch 78/500
27/27 [==============================] - 0s 882us/step - loss: 0.7866 - sparse_categorical_accuracy: 0.7651
Epoch 79/500
27/27 [==============================] - 0s 891us/step - loss: 0.7754 - sparse_categorical_accuracy: 0.7627
Epoch 80/500
27/27 [==============================] - 0s 911us/step - loss: 0.7940 - sparse_categorical_accuracy: 0.7627
Epoch 81/500
27/27 [==============================] - 0s 957us/step - loss: 0.7881 - sparse_categorical_accuracy: 0.7603
Epoch 82/500
27/27 [==============================] - 0s 1ms/step - loss: 0.7544 - sparse_categorical_accuracy: 0.7757
Epoch 83/500
27/27 [==============================] - 0s 931us/step - loss: 0.7647 - sparse_categorical_accuracy: 0.7733
Epoch 84/500
27/27 [==============================] - 0s 964us/step - loss: 0.7464 - sparse_categorical_accuracy: 0.7863
Epoch 85/500
27/27 [==============================] - 0s 895us/step - loss: 0.8166 - sparse_categorical_accuracy: 0.7273
Epoch 86/500
27/27 [==============================] - 0s 888us/step - loss: 0.7509 - sparse_categorical_accuracy: 0.7839
Epoch 87/500
27/27 [==============================] - 0s 891us/step - loss: 0.7769 - sparse_categorical_accuracy: 0.7438
Epoch 88/500
27/27 [==============================] - 0s 951us/step - loss: 0.7473 - sparse_categorical_accuracy: 0.7804
Epoch 89/500
27/27 [==============================] - 0s 988us/step - loss: 0.7138 - sparse_categorical_accuracy: 0.8111
Epoch 90/500
27/27 [==============================] - 0s 918us/step - loss: 0.7360 - sparse_categorical_accuracy: 0.7910
Epoch 91/500
27/27 [==============================] - 0s 904us/step - loss: 0.7179 - sparse_categorical_accuracy: 0.7946
Epoch 92/500
27/27 [==============================] - 0s 908us/step - loss: 0.6990 - sparse_categorical_accuracy: 0.8288
Epoch 93/500
27/27 [==============================] - 0s 925us/step - loss: 0.7166 - sparse_categorical_accuracy: 0.8064
Epoch 94/500
27/27 [==============================] - 0s 913us/step - loss: 0.7004 - sparse_categorical_accuracy: 0.8040
Epoch 95/500
27/27 [==============================] - 0s 885us/step - loss: 0.7578 - sparse_categorical_accuracy: 0.7651
Epoch 96/500
27/27 [==============================] - 0s 903us/step - loss: 0.7159 - sparse_categorical_accuracy: 0.8005
Epoch 97/500
27/27 [==============================] - 0s 995us/step - loss: 0.6972 - sparse_categorical_accuracy: 0.8135
Epoch 98/500
27/27 [==============================] - 0s 906us/step - loss: 0.7073 - sparse_categorical_accuracy: 0.8087
Epoch 99/500
27/27 [==============================] - 0s 893us/step - loss: 0.7451 - sparse_categorical_accuracy: 0.7780
Epoch 100/500
27/27 [==============================] - 0s 5ms/step - loss: 0.6952 - sparse_categorical_accuracy: 0.8099 - val_loss: 0.9220 - val_sparse_categorical_accuracy: 0.7217
Epoch 101/500
27/27 [==============================] - 0s 869us/step - loss: 0.7454 - sparse_categorical_accuracy: 0.7863
Epoch 102/500
27/27 [==============================] - 0s 921us/step - loss: 0.7431 - sparse_categorical_accuracy: 0.7934
Epoch 103/500
27/27 [==============================] - 0s 934us/step - loss: 0.6931 - sparse_categorical_accuracy: 0.8005
Epoch 104/500
27/27 [==============================] - 0s 1ms/step - loss: 0.6887 - sparse_categorical_accuracy: 0.8064
Epoch 105/500
27/27 [==============================] - 0s 1ms/step - loss: 0.6730 - sparse_categorical_accuracy: 0.8170
Epoch 106/500
27/27 [==============================] - 0s 1ms/step - loss: 0.7147 - sparse_categorical_accuracy: 0.8040
Epoch 107/500
27/27 [==============================] - 0s 925us/step - loss: 0.7189 - sparse_categorical_accuracy: 0.7910
Epoch 108/500
27/27 [==============================] - 0s 963us/step - loss: 0.6830 - sparse_categorical_accuracy: 0.8111
Epoch 109/500
27/27 [==============================] - 0s 1ms/step - loss: 0.6995 - sparse_categorical_accuracy: 0.8158
Epoch 110/500
27/27 [==============================] - 0s 887us/step - loss: 0.6874 - sparse_categorical_accuracy: 0.8052
Epoch 111/500
27/27 [==============================] - 0s 921us/step - loss: 0.6595 - sparse_categorical_accuracy: 0.8276
Epoch 112/500
27/27 [==============================] - 0s 1ms/step - loss: 0.7001 - sparse_categorical_accuracy: 0.7910
Epoch 113/500
27/27 [==============================] - 0s 906us/step - loss: 0.6490 - sparse_categorical_accuracy: 0.8264
Epoch 114/500
27/27 [==============================] - 0s 829us/step - loss: 0.6860 - sparse_categorical_accuracy: 0.8076
Epoch 115/500
27/27 [==============================] - 0s 939us/step - loss: 0.6636 - sparse_categorical_accuracy: 0.8158
Epoch 116/500
27/27 [==============================] - 0s 917us/step - loss: 0.6662 - sparse_categorical_accuracy: 0.8241
Epoch 117/500
27/27 [==============================] - 0s 914us/step - loss: 0.6623 - sparse_categorical_accuracy: 0.8205
Epoch 118/500
27/27 [==============================] - 0s 881us/step - loss: 0.6576 - sparse_categorical_accuracy: 0.8276
Epoch 119/500
27/27 [==============================] - 0s 924us/step - loss: 0.6537 - sparse_categorical_accuracy: 0.8123
Epoch 120/500
27/27 [==============================] - 0s 911us/step - loss: 0.6368 - sparse_categorical_accuracy: 0.8264
Epoch 121/500
27/27 [==============================] - 0s 961us/step - loss: 0.6388 - sparse_categorical_accuracy: 0.8312
Epoch 122/500
27/27 [==============================] - 0s 946us/step - loss: 0.6272 - sparse_categorical_accuracy: 0.8229
Epoch 123/500
27/27 [==============================] - 0s 848us/step - loss: 0.6375 - sparse_categorical_accuracy: 0.8205
Epoch 124/500
27/27 [==============================] - 0s 924us/step - loss: 0.6118 - sparse_categorical_accuracy: 0.8430
Epoch 125/500
27/27 [==============================] - 0s 877us/step - loss: 0.6168 - sparse_categorical_accuracy: 0.8253
Epoch 126/500
27/27 [==============================] - 0s 921us/step - loss: 0.6371 - sparse_categorical_accuracy: 0.8264
Epoch 127/500
27/27 [==============================] - 0s 941us/step - loss: 0.6289 - sparse_categorical_accuracy: 0.8241
Epoch 128/500
27/27 [==============================] - 0s 897us/step - loss: 0.6146 - sparse_categorical_accuracy: 0.8288
Epoch 129/500
27/27 [==============================] - 0s 925us/step - loss: 0.6426 - sparse_categorical_accuracy: 0.8182
Epoch 130/500
27/27 [==============================] - 0s 922us/step - loss: 0.6036 - sparse_categorical_accuracy: 0.8654
Epoch 131/500
27/27 [==============================] - 0s 891us/step - loss: 0.6938 - sparse_categorical_accuracy: 0.8028
Epoch 132/500
27/27 [==============================] - 0s 930us/step - loss: 0.6204 - sparse_categorical_accuracy: 0.8548
Epoch 133/500
27/27 [==============================] - 0s 952us/step - loss: 0.6165 - sparse_categorical_accuracy: 0.8583
Epoch 134/500
27/27 [==============================] - 0s 877us/step - loss: 0.6194 - sparse_categorical_accuracy: 0.8524
Epoch 135/500
27/27 [==============================] - 0s 936us/step - loss: 0.6111 - sparse_categorical_accuracy: 0.8630
Epoch 136/500
27/27 [==============================] - 0s 829us/step - loss: 0.5766 - sparse_categorical_accuracy: 0.8855
Epoch 137/500
27/27 [==============================] - 0s 948us/step - loss: 0.5779 - sparse_categorical_accuracy: 0.8678
Epoch 138/500
27/27 [==============================] - 0s 828us/step - loss: 0.5750 - sparse_categorical_accuracy: 0.8819
Epoch 139/500
27/27 [==============================] - 0s 919us/step - loss: 0.5946 - sparse_categorical_accuracy: 0.8796
Epoch 140/500
27/27 [==============================] - 0s 919us/step - loss: 0.5693 - sparse_categorical_accuracy: 0.8890
Epoch 141/500
27/27 [==============================] - 0s 910us/step - loss: 0.5717 - sparse_categorical_accuracy: 0.8914
Epoch 142/500
27/27 [==============================] - 0s 966us/step - loss: 0.5723 - sparse_categorical_accuracy: 0.8878
Epoch 143/500
27/27 [==============================] - 0s 883us/step - loss: 0.5872 - sparse_categorical_accuracy: 0.8784
Epoch 144/500
27/27 [==============================] - 0s 846us/step - loss: 0.5940 - sparse_categorical_accuracy: 0.8725
Epoch 145/500
27/27 [==============================] - 0s 828us/step - loss: 0.5727 - sparse_categorical_accuracy: 0.8890
Epoch 146/500
27/27 [==============================] - 0s 861us/step - loss: 0.5764 - sparse_categorical_accuracy: 0.8878
Epoch 147/500
27/27 [==============================] - 0s 875us/step - loss: 0.5413 - sparse_categorical_accuracy: 0.9032
Epoch 148/500
27/27 [==============================] - 0s 832us/step - loss: 0.5615 - sparse_categorical_accuracy: 0.8855
Epoch 149/500
27/27 [==============================] - 0s 920us/step - loss: 0.5481 - sparse_categorical_accuracy: 0.9055
Epoch 150/500
27/27 [==============================] - 0s 961us/step - loss: 0.5751 - sparse_categorical_accuracy: 0.8914
Epoch 151/500
27/27 [==============================] - 0s 949us/step - loss: 0.5198 - sparse_categorical_accuracy: 0.9055
Epoch 152/500
27/27 [==============================] - 0s 909us/step - loss: 0.5125 - sparse_categorical_accuracy: 0.9126
Epoch 153/500
27/27 [==============================] - 0s 903us/step - loss: 0.5143 - sparse_categorical_accuracy: 0.9221
Epoch 154/500
27/27 [==============================] - 0s 906us/step - loss: 0.5842 - sparse_categorical_accuracy: 0.8678
Epoch 155/500
27/27 [==============================] - 0s 868us/step - loss: 0.5322 - sparse_categorical_accuracy: 0.9044
Epoch 156/500
27/27 [==============================] - 0s 832us/step - loss: 0.5105 - sparse_categorical_accuracy: 0.9162
Epoch 157/500
27/27 [==============================] - 0s 866us/step - loss: 0.5454 - sparse_categorical_accuracy: 0.8902
Epoch 158/500
27/27 [==============================] - 0s 1ms/step - loss: 0.5075 - sparse_categorical_accuracy: 0.9115
Epoch 159/500
27/27 [==============================] - 0s 966us/step - loss: 0.5293 - sparse_categorical_accuracy: 0.9032
Epoch 160/500
27/27 [==============================] - 0s 928us/step - loss: 0.5273 - sparse_categorical_accuracy: 0.9020
Epoch 161/500
27/27 [==============================] - 0s 860us/step - loss: 0.4955 - sparse_categorical_accuracy: 0.9174
Epoch 162/500
27/27 [==============================] - 0s 902us/step - loss: 0.4818 - sparse_categorical_accuracy: 0.9197
Epoch 163/500
27/27 [==============================] - 0s 847us/step - loss: 0.5052 - sparse_categorical_accuracy: 0.9020
Epoch 164/500
27/27 [==============================] - 0s 966us/step - loss: 0.5142 - sparse_categorical_accuracy: 0.9044
Epoch 165/500
27/27 [==============================] - 0s 849us/step - loss: 0.5139 - sparse_categorical_accuracy: 0.9126
Epoch 166/500
27/27 [==============================] - 0s 805us/step - loss: 0.5285 - sparse_categorical_accuracy: 0.8985
Epoch 167/500
27/27 [==============================] - 0s 879us/step - loss: 0.4723 - sparse_categorical_accuracy: 0.9268
Epoch 168/500
27/27 [==============================] - 0s 917us/step - loss: 0.4844 - sparse_categorical_accuracy: 0.9067
Epoch 169/500
27/27 [==============================] - 0s 886us/step - loss: 0.4956 - sparse_categorical_accuracy: 0.9103
Epoch 170/500
27/27 [==============================] - 0s 926us/step - loss: 0.4799 - sparse_categorical_accuracy: 0.9138
Epoch 171/500
27/27 [==============================] - 0s 949us/step - loss: 0.5113 - sparse_categorical_accuracy: 0.9079
Epoch 172/500
27/27 [==============================] - 0s 858us/step - loss: 0.4909 - sparse_categorical_accuracy: 0.9067
Epoch 173/500
27/27 [==============================] - 0s 908us/step - loss: 0.4714 - sparse_categorical_accuracy: 0.9103
Epoch 174/500
27/27 [==============================] - 0s 905us/step - loss: 0.4913 - sparse_categorical_accuracy: 0.9138
Epoch 175/500
27/27 [==============================] - 0s 838us/step - loss: 0.4989 - sparse_categorical_accuracy: 0.9103
Epoch 176/500
27/27 [==============================] - 0s 901us/step - loss: 0.4738 - sparse_categorical_accuracy: 0.9221
Epoch 177/500
27/27 [==============================] - 0s 884us/step - loss: 0.4929 - sparse_categorical_accuracy: 0.9032
Epoch 178/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4433 - sparse_categorical_accuracy: 0.9303
Epoch 179/500
27/27 [==============================] - 0s 901us/step - loss: 0.4551 - sparse_categorical_accuracy: 0.9303
Epoch 180/500
27/27 [==============================] - 0s 870us/step - loss: 0.4909 - sparse_categorical_accuracy: 0.9020
Epoch 181/500
27/27 [==============================] - 0s 924us/step - loss: 0.4522 - sparse_categorical_accuracy: 0.9185
Epoch 182/500
27/27 [==============================] - 0s 893us/step - loss: 0.4479 - sparse_categorical_accuracy: 0.9303
Epoch 183/500
27/27 [==============================] - 0s 904us/step - loss: 0.4413 - sparse_categorical_accuracy: 0.9256
Epoch 184/500
27/27 [==============================] - 0s 899us/step - loss: 0.4498 - sparse_categorical_accuracy: 0.9174
Epoch 185/500
27/27 [==============================] - 0s 843us/step - loss: 0.4517 - sparse_categorical_accuracy: 0.9162
Epoch 186/500
27/27 [==============================] - 0s 931us/step - loss: 0.4478 - sparse_categorical_accuracy: 0.9292
Epoch 187/500
27/27 [==============================] - 0s 846us/step - loss: 0.4498 - sparse_categorical_accuracy: 0.9138
Epoch 188/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4591 - sparse_categorical_accuracy: 0.9162
Epoch 189/500
27/27 [==============================] - 0s 878us/step - loss: 0.4742 - sparse_categorical_accuracy: 0.9044
Epoch 190/500
27/27 [==============================] - 0s 894us/step - loss: 0.4385 - sparse_categorical_accuracy: 0.9315
Epoch 191/500
27/27 [==============================] - 0s 920us/step - loss: 0.4501 - sparse_categorical_accuracy: 0.9280
Epoch 192/500
27/27 [==============================] - 0s 788us/step - loss: 0.4786 - sparse_categorical_accuracy: 0.9115
Epoch 193/500
27/27 [==============================] - 0s 932us/step - loss: 0.8577 - sparse_categorical_accuracy: 0.7568
Epoch 194/500
27/27 [==============================] - 0s 819us/step - loss: 0.7656 - sparse_categorical_accuracy: 0.7839
Epoch 195/500
27/27 [==============================] - 0s 957us/step - loss: 0.6410 - sparse_categorical_accuracy: 0.8453
Epoch 196/500
27/27 [==============================] - 0s 915us/step - loss: 0.6841 - sparse_categorical_accuracy: 0.8241
Epoch 197/500
27/27 [==============================] - 0s 876us/step - loss: 0.6684 - sparse_categorical_accuracy: 0.8241
Epoch 198/500
27/27 [==============================] - 0s 953us/step - loss: 0.6583 - sparse_categorical_accuracy: 0.8312
Epoch 199/500
27/27 [==============================] - 0s 826us/step - loss: 0.5459 - sparse_categorical_accuracy: 0.8843
Epoch 200/500
27/27 [==============================] - 0s 2ms/step - loss: 0.5818 - sparse_categorical_accuracy: 0.8843 - val_loss: 0.6180 - val_sparse_categorical_accuracy: 0.8302
Epoch 201/500
27/27 [==============================] - 0s 889us/step - loss: 0.5705 - sparse_categorical_accuracy: 0.8843
Epoch 202/500
27/27 [==============================] - 0s 872us/step - loss: 0.6053 - sparse_categorical_accuracy: 0.8548
Epoch 203/500
27/27 [==============================] - 0s 916us/step - loss: 0.6325 - sparse_categorical_accuracy: 0.8442
Epoch 204/500
27/27 [==============================] - 0s 828us/step - loss: 0.5133 - sparse_categorical_accuracy: 0.9008
Epoch 205/500
27/27 [==============================] - 0s 954us/step - loss: 0.6436 - sparse_categorical_accuracy: 0.8501
Epoch 206/500
27/27 [==============================] - 0s 878us/step - loss: 0.4927 - sparse_categorical_accuracy: 0.9268
Epoch 207/500
27/27 [==============================] - 0s 994us/step - loss: 0.4982 - sparse_categorical_accuracy: 0.9126
Epoch 208/500
27/27 [==============================] - 0s 860us/step - loss: 0.5219 - sparse_categorical_accuracy: 0.9044
Epoch 209/500
27/27 [==============================] - 0s 894us/step - loss: 0.5310 - sparse_categorical_accuracy: 0.8996
Epoch 210/500
27/27 [==============================] - 0s 884us/step - loss: 0.5166 - sparse_categorical_accuracy: 0.9103
Epoch 211/500
27/27 [==============================] - 0s 922us/step - loss: 0.5291 - sparse_categorical_accuracy: 0.8843
Epoch 212/500
27/27 [==============================] - 0s 974us/step - loss: 0.4659 - sparse_categorical_accuracy: 0.9221
Epoch 213/500
27/27 [==============================] - 0s 887us/step - loss: 0.4971 - sparse_categorical_accuracy: 0.9209
Epoch 214/500
27/27 [==============================] - 0s 958us/step - loss: 0.4663 - sparse_categorical_accuracy: 0.9351
Epoch 215/500
27/27 [==============================] - 0s 852us/step - loss: 0.4781 - sparse_categorical_accuracy: 0.9091
Epoch 216/500
27/27 [==============================] - 0s 854us/step - loss: 0.8753 - sparse_categorical_accuracy: 0.7710
Epoch 217/500
27/27 [==============================] - 0s 921us/step - loss: 0.6263 - sparse_categorical_accuracy: 0.8465
Epoch 218/500
27/27 [==============================] - 0s 846us/step - loss: 0.5606 - sparse_categorical_accuracy: 0.8926
Epoch 219/500
27/27 [==============================] - 0s 970us/step - loss: 0.5323 - sparse_categorical_accuracy: 0.9020
Epoch 220/500
27/27 [==============================] - 0s 911us/step - loss: 0.5313 - sparse_categorical_accuracy: 0.9162
Epoch 221/500
27/27 [==============================] - 0s 898us/step - loss: 0.5543 - sparse_categorical_accuracy: 0.8949
Epoch 222/500
27/27 [==============================] - 0s 900us/step - loss: 0.5170 - sparse_categorical_accuracy: 0.9138
Epoch 223/500
27/27 [==============================] - 0s 891us/step - loss: 0.5134 - sparse_categorical_accuracy: 0.9091
Epoch 224/500
27/27 [==============================] - 0s 969us/step - loss: 0.4835 - sparse_categorical_accuracy: 0.9303
Epoch 225/500
27/27 [==============================] - 0s 915us/step - loss: 0.5224 - sparse_categorical_accuracy: 0.9150
Epoch 226/500
27/27 [==============================] - 0s 946us/step - loss: 0.5245 - sparse_categorical_accuracy: 0.9115
Epoch 227/500
27/27 [==============================] - 0s 903us/step - loss: 0.4805 - sparse_categorical_accuracy: 0.9256
Epoch 228/500
27/27 [==============================] - 0s 883us/step - loss: 0.4679 - sparse_categorical_accuracy: 0.9445
Epoch 229/500
27/27 [==============================] - 0s 922us/step - loss: 0.4859 - sparse_categorical_accuracy: 0.9150
Epoch 230/500
27/27 [==============================] - 0s 872us/step - loss: 0.4979 - sparse_categorical_accuracy: 0.9280
Epoch 231/500
27/27 [==============================] - 0s 829us/step - loss: 0.5351 - sparse_categorical_accuracy: 0.8949
Epoch 232/500
27/27 [==============================] - 0s 849us/step - loss: 0.4932 - sparse_categorical_accuracy: 0.9150
Epoch 233/500
27/27 [==============================] - 0s 905us/step - loss: 0.4898 - sparse_categorical_accuracy: 0.9221
Epoch 234/500
27/27 [==============================] - 0s 897us/step - loss: 0.4663 - sparse_categorical_accuracy: 0.9292
Epoch 235/500
27/27 [==============================] - 0s 815us/step - loss: 0.4764 - sparse_categorical_accuracy: 0.9315
Epoch 236/500
27/27 [==============================] - 0s 891us/step - loss: 0.4716 - sparse_categorical_accuracy: 0.9268
Epoch 237/500
27/27 [==============================] - 0s 851us/step - loss: 0.4625 - sparse_categorical_accuracy: 0.9280
Epoch 238/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4851 - sparse_categorical_accuracy: 0.9150
Epoch 239/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4528 - sparse_categorical_accuracy: 0.9421
Epoch 240/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4559 - sparse_categorical_accuracy: 0.9351
Epoch 241/500
27/27 [==============================] - 0s 956us/step - loss: 0.4689 - sparse_categorical_accuracy: 0.9256
Epoch 242/500
27/27 [==============================] - 0s 914us/step - loss: 0.4644 - sparse_categorical_accuracy: 0.9185
Epoch 243/500
27/27 [==============================] - 0s 978us/step - loss: 0.4662 - sparse_categorical_accuracy: 0.9268
Epoch 244/500
27/27 [==============================] - 0s 922us/step - loss: 0.4631 - sparse_categorical_accuracy: 0.9327
Epoch 245/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4798 - sparse_categorical_accuracy: 0.9174
Epoch 246/500
27/27 [==============================] - 0s 894us/step - loss: 0.4831 - sparse_categorical_accuracy: 0.9115
Epoch 247/500
27/27 [==============================] - 0s 866us/step - loss: 0.4492 - sparse_categorical_accuracy: 0.9303
Epoch 248/500
27/27 [==============================] - 0s 891us/step - loss: 0.4465 - sparse_categorical_accuracy: 0.9221
Epoch 249/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4688 - sparse_categorical_accuracy: 0.9185
Epoch 250/500
27/27 [==============================] - 0s 922us/step - loss: 0.4528 - sparse_categorical_accuracy: 0.9115
Epoch 251/500
27/27 [==============================] - 0s 870us/step - loss: 0.4545 - sparse_categorical_accuracy: 0.9233
Epoch 252/500
27/27 [==============================] - 0s 912us/step - loss: 0.4520 - sparse_categorical_accuracy: 0.9209
Epoch 253/500
27/27 [==============================] - 0s 921us/step - loss: 0.4407 - sparse_categorical_accuracy: 0.9244
Epoch 254/500
27/27 [==============================] - 0s 882us/step - loss: 0.4642 - sparse_categorical_accuracy: 0.9162
Epoch 255/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4712 - sparse_categorical_accuracy: 0.9197
Epoch 256/500
27/27 [==============================] - 0s 920us/step - loss: 0.4355 - sparse_categorical_accuracy: 0.9292
Epoch 257/500
27/27 [==============================] - 0s 896us/step - loss: 0.4480 - sparse_categorical_accuracy: 0.9280
Epoch 258/500
27/27 [==============================] - 0s 894us/step - loss: 0.4551 - sparse_categorical_accuracy: 0.9185
Epoch 259/500
27/27 [==============================] - 0s 930us/step - loss: 0.4265 - sparse_categorical_accuracy: 0.9303
Epoch 260/500
27/27 [==============================] - 0s 955us/step - loss: 0.4351 - sparse_categorical_accuracy: 0.9303
Epoch 261/500
27/27 [==============================] - 0s 907us/step - loss: 0.4446 - sparse_categorical_accuracy: 0.9185
Epoch 262/500
27/27 [==============================] - 0s 921us/step - loss: 0.4341 - sparse_categorical_accuracy: 0.9185
Epoch 263/500
27/27 [==============================] - 0s 902us/step - loss: 0.4323 - sparse_categorical_accuracy: 0.9244
Epoch 264/500
27/27 [==============================] - 0s 880us/step - loss: 0.4646 - sparse_categorical_accuracy: 0.9126
Epoch 265/500
27/27 [==============================] - 0s 934us/step - loss: 0.4455 - sparse_categorical_accuracy: 0.9268
Epoch 266/500
27/27 [==============================] - 0s 999us/step - loss: 0.4367 - sparse_categorical_accuracy: 0.9315
Epoch 267/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4499 - sparse_categorical_accuracy: 0.9221
Epoch 268/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4370 - sparse_categorical_accuracy: 0.9268
Epoch 269/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4514 - sparse_categorical_accuracy: 0.9185
Epoch 270/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4524 - sparse_categorical_accuracy: 0.9197
Epoch 271/500
27/27 [==============================] - 0s 964us/step - loss: 0.4208 - sparse_categorical_accuracy: 0.9197
Epoch 272/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4359 - sparse_categorical_accuracy: 0.9221
Epoch 273/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4430 - sparse_categorical_accuracy: 0.9091
Epoch 274/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4331 - sparse_categorical_accuracy: 0.9292
Epoch 275/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4371 - sparse_categorical_accuracy: 0.9303
Epoch 276/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4251 - sparse_categorical_accuracy: 0.9292
Epoch 277/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4709 - sparse_categorical_accuracy: 0.9091
Epoch 278/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4422 - sparse_categorical_accuracy: 0.9233
Epoch 279/500
27/27 [==============================] - 0s 939us/step - loss: 0.4343 - sparse_categorical_accuracy: 0.9233
Epoch 280/500
27/27 [==============================] - 0s 983us/step - loss: 0.4235 - sparse_categorical_accuracy: 0.9174
Epoch 281/500
27/27 [==============================] - 0s 998us/step - loss: 0.4172 - sparse_categorical_accuracy: 0.9209
Epoch 282/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4209 - sparse_categorical_accuracy: 0.9327
Epoch 283/500
27/27 [==============================] - 0s 889us/step - loss: 0.4193 - sparse_categorical_accuracy: 0.9327
Epoch 284/500
27/27 [==============================] - 0s 966us/step - loss: 0.4276 - sparse_categorical_accuracy: 0.9315
Epoch 285/500
27/27 [==============================] - 0s 957us/step - loss: 0.4333 - sparse_categorical_accuracy: 0.9185
Epoch 286/500
27/27 [==============================] - 0s 959us/step - loss: 0.4231 - sparse_categorical_accuracy: 0.9280
Epoch 287/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4238 - sparse_categorical_accuracy: 0.9303
Epoch 288/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4168 - sparse_categorical_accuracy: 0.9327
Epoch 289/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4387 - sparse_categorical_accuracy: 0.9256
Epoch 290/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4315 - sparse_categorical_accuracy: 0.9197
Epoch 291/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4421 - sparse_categorical_accuracy: 0.9174
Epoch 292/500
27/27 [==============================] - 0s 997us/step - loss: 0.4468 - sparse_categorical_accuracy: 0.9185
Epoch 293/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4290 - sparse_categorical_accuracy: 0.9244
Epoch 294/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4279 - sparse_categorical_accuracy: 0.9292
Epoch 295/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4298 - sparse_categorical_accuracy: 0.9197
Epoch 296/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4276 - sparse_categorical_accuracy: 0.9256
Epoch 297/500
27/27 [==============================] - 0s 984us/step - loss: 0.4504 - sparse_categorical_accuracy: 0.9126
Epoch 298/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4267 - sparse_categorical_accuracy: 0.9268
Epoch 299/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4237 - sparse_categorical_accuracy: 0.9268
Epoch 300/500
27/27 [==============================] - 0s 2ms/step - loss: 0.4234 - sparse_categorical_accuracy: 0.9197 - val_loss: 0.4862 - val_sparse_categorical_accuracy: 0.9009
Epoch 301/500
27/27 [==============================] - 0s 942us/step - loss: 0.4251 - sparse_categorical_accuracy: 0.9280
Epoch 302/500
27/27 [==============================] - 0s 815us/step - loss: 0.4145 - sparse_categorical_accuracy: 0.9280
Epoch 303/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4209 - sparse_categorical_accuracy: 0.9256
Epoch 304/500
27/27 [==============================] - 0s 941us/step - loss: 0.4181 - sparse_categorical_accuracy: 0.9339
Epoch 305/500
27/27 [==============================] - 0s 913us/step - loss: 0.4181 - sparse_categorical_accuracy: 0.9292
Epoch 306/500
27/27 [==============================] - 0s 936us/step - loss: 0.4263 - sparse_categorical_accuracy: 0.9221
Epoch 307/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4386 - sparse_categorical_accuracy: 0.9221
Epoch 308/500
27/27 [==============================] - 0s 927us/step - loss: 0.4195 - sparse_categorical_accuracy: 0.9327
Epoch 309/500
27/27 [==============================] - 0s 991us/step - loss: 0.4153 - sparse_categorical_accuracy: 0.9233
Epoch 310/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4197 - sparse_categorical_accuracy: 0.9268
Epoch 311/500
27/27 [==============================] - 0s 987us/step - loss: 0.4112 - sparse_categorical_accuracy: 0.9280
Epoch 312/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4134 - sparse_categorical_accuracy: 0.9244
Epoch 313/500
27/27 [==============================] - 0s 943us/step - loss: 0.4155 - sparse_categorical_accuracy: 0.9303
Epoch 314/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4146 - sparse_categorical_accuracy: 0.9280
Epoch 315/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4296 - sparse_categorical_accuracy: 0.9209
Epoch 316/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4223 - sparse_categorical_accuracy: 0.9256
Epoch 317/500
27/27 [==============================] - 0s 919us/step - loss: 0.4143 - sparse_categorical_accuracy: 0.9233
Epoch 318/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4137 - sparse_categorical_accuracy: 0.9256
Epoch 319/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4136 - sparse_categorical_accuracy: 0.9221
Epoch 320/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4236 - sparse_categorical_accuracy: 0.9209
Epoch 321/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4189 - sparse_categorical_accuracy: 0.9162
Epoch 322/500
27/27 [==============================] - 0s 966us/step - loss: 0.4292 - sparse_categorical_accuracy: 0.9244
Epoch 323/500
27/27 [==============================] - 0s 900us/step - loss: 0.4261 - sparse_categorical_accuracy: 0.9126
Epoch 324/500
27/27 [==============================] - 0s 1ms/step - loss: 0.3953 - sparse_categorical_accuracy: 0.9315
Epoch 325/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4123 - sparse_categorical_accuracy: 0.9256
Epoch 326/500
27/27 [==============================] - 0s 921us/step - loss: 0.4052 - sparse_categorical_accuracy: 0.9327
Epoch 327/500
27/27 [==============================] - 0s 951us/step - loss: 0.3970 - sparse_categorical_accuracy: 0.9256
Epoch 328/500
27/27 [==============================] - 0s 992us/step - loss: 0.4007 - sparse_categorical_accuracy: 0.9339
Epoch 329/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4056 - sparse_categorical_accuracy: 0.9174
Epoch 330/500
27/27 [==============================] - 0s 1ms/step - loss: 0.3985 - sparse_categorical_accuracy: 0.9315
Epoch 331/500
27/27 [==============================] - 0s 924us/step - loss: 0.3995 - sparse_categorical_accuracy: 0.9303
Epoch 332/500
27/27 [==============================] - 0s 951us/step - loss: 0.4010 - sparse_categorical_accuracy: 0.9256
Epoch 333/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4118 - sparse_categorical_accuracy: 0.9174
Epoch 334/500
27/27 [==============================] - 0s 994us/step - loss: 0.4047 - sparse_categorical_accuracy: 0.9268
Epoch 335/500
27/27 [==============================] - 0s 968us/step - loss: 0.3990 - sparse_categorical_accuracy: 0.9268
Epoch 336/500
27/27 [==============================] - 0s 992us/step - loss: 0.4063 - sparse_categorical_accuracy: 0.9221
Epoch 337/500
27/27 [==============================] - 0s 994us/step - loss: 0.3987 - sparse_categorical_accuracy: 0.9362
Epoch 338/500
27/27 [==============================] - 0s 1ms/step - loss: 0.3977 - sparse_categorical_accuracy: 0.9339
Epoch 339/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4041 - sparse_categorical_accuracy: 0.9244
Epoch 340/500
27/27 [==============================] - 0s 987us/step - loss: 0.3989 - sparse_categorical_accuracy: 0.9280
Epoch 341/500
27/27 [==============================] - 0s 890us/step - loss: 0.3923 - sparse_categorical_accuracy: 0.9315
Epoch 342/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4158 - sparse_categorical_accuracy: 0.9221
Epoch 343/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4068 - sparse_categorical_accuracy: 0.9244
Epoch 344/500
27/27 [==============================] - 0s 1ms/step - loss: 0.3938 - sparse_categorical_accuracy: 0.9268
Epoch 345/500
27/27 [==============================] - 0s 918us/step - loss: 0.3955 - sparse_categorical_accuracy: 0.9221
Epoch 346/500
27/27 [==============================] - 0s 1ms/step - loss: 0.3933 - sparse_categorical_accuracy: 0.9339
Epoch 347/500
27/27 [==============================] - 0s 1ms/step - loss: 0.3979 - sparse_categorical_accuracy: 0.9292
Epoch 348/500
27/27 [==============================] - 0s 964us/step - loss: 0.4074 - sparse_categorical_accuracy: 0.9303
Epoch 349/500
27/27 [==============================] - 0s 916us/step - loss: 0.4044 - sparse_categorical_accuracy: 0.9268
Epoch 350/500
27/27 [==============================] - 0s 953us/step - loss: 0.3896 - sparse_categorical_accuracy: 0.9445
Epoch 351/500
27/27 [==============================] - 0s 1ms/step - loss: 0.3912 - sparse_categorical_accuracy: 0.9268
Epoch 352/500
27/27 [==============================] - 0s 995us/step - loss: 0.4089 - sparse_categorical_accuracy: 0.9174
Epoch 353/500
27/27 [==============================] - 0s 919us/step - loss: 0.3987 - sparse_categorical_accuracy: 0.9292
Epoch 354/500
27/27 [==============================] - 0s 992us/step - loss: 0.4009 - sparse_categorical_accuracy: 0.9256
Epoch 355/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4162 - sparse_categorical_accuracy: 0.9055
Epoch 356/500
27/27 [==============================] - 0s 960us/step - loss: 0.4001 - sparse_categorical_accuracy: 0.9280
Epoch 357/500
27/27 [==============================] - 0s 854us/step - loss: 0.4024 - sparse_categorical_accuracy: 0.9244
Epoch 358/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4005 - sparse_categorical_accuracy: 0.9339
Epoch 359/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4022 - sparse_categorical_accuracy: 0.9280
Epoch 360/500
27/27 [==============================] - 0s 960us/step - loss: 0.3940 - sparse_categorical_accuracy: 0.9221
Epoch 361/500
27/27 [==============================] - 0s 913us/step - loss: 0.3889 - sparse_categorical_accuracy: 0.9315
Epoch 362/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4115 - sparse_categorical_accuracy: 0.9197
Epoch 363/500
27/27 [==============================] - 0s 1ms/step - loss: 0.3932 - sparse_categorical_accuracy: 0.9280
Epoch 364/500
27/27 [==============================] - 0s 895us/step - loss: 0.4010 - sparse_categorical_accuracy: 0.9268
Epoch 365/500
27/27 [==============================] - 0s 842us/step - loss: 0.3948 - sparse_categorical_accuracy: 0.9433
Epoch 366/500
27/27 [==============================] - 0s 961us/step - loss: 0.4089 - sparse_categorical_accuracy: 0.9303
Epoch 367/500
27/27 [==============================] - 0s 916us/step - loss: 0.4158 - sparse_categorical_accuracy: 0.9233
Epoch 368/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4151 - sparse_categorical_accuracy: 0.9233
Epoch 369/500
27/27 [==============================] - 0s 997us/step - loss: 0.3966 - sparse_categorical_accuracy: 0.9303
Epoch 370/500
27/27 [==============================] - 0s 987us/step - loss: 0.3907 - sparse_categorical_accuracy: 0.9280
Epoch 371/500
27/27 [==============================] - 0s 876us/step - loss: 0.3963 - sparse_categorical_accuracy: 0.9280
Epoch 372/500
27/27 [==============================] - 0s 950us/step - loss: 0.4117 - sparse_categorical_accuracy: 0.9162
Epoch 373/500
27/27 [==============================] - 0s 1ms/step - loss: 0.3875 - sparse_categorical_accuracy: 0.9327
Epoch 374/500
27/27 [==============================] - 0s 1ms/step - loss: 0.3900 - sparse_categorical_accuracy: 0.9256
Epoch 375/500
27/27 [==============================] - 0s 1ms/step - loss: 0.3902 - sparse_categorical_accuracy: 0.9362
Epoch 376/500
27/27 [==============================] - 0s 1ms/step - loss: 0.3970 - sparse_categorical_accuracy: 0.9292
Epoch 377/500
27/27 [==============================] - 0s 1ms/step - loss: 0.3947 - sparse_categorical_accuracy: 0.9327
Epoch 378/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4000 - sparse_categorical_accuracy: 0.9280
Epoch 379/500
27/27 [==============================] - 0s 1ms/step - loss: 0.3951 - sparse_categorical_accuracy: 0.9327
Epoch 380/500
27/27 [==============================] - 0s 998us/step - loss: 0.4030 - sparse_categorical_accuracy: 0.9162
Epoch 381/500
27/27 [==============================] - 0s 933us/step - loss: 0.3934 - sparse_categorical_accuracy: 0.9339
Epoch 382/500
27/27 [==============================] - 0s 976us/step - loss: 0.4039 - sparse_categorical_accuracy: 0.9174
Epoch 383/500
27/27 [==============================] - 0s 1ms/step - loss: 0.3963 - sparse_categorical_accuracy: 0.9292
Epoch 384/500
27/27 [==============================] - 0s 892us/step - loss: 0.3821 - sparse_categorical_accuracy: 0.9303
Epoch 385/500
27/27 [==============================] - 0s 870us/step - loss: 0.3921 - sparse_categorical_accuracy: 0.9339
Epoch 386/500
27/27 [==============================] - 0s 1ms/step - loss: 0.3856 - sparse_categorical_accuracy: 0.9327
Epoch 387/500
27/27 [==============================] - 0s 955us/step - loss: 0.4011 - sparse_categorical_accuracy: 0.9256
Epoch 388/500
27/27 [==============================] - 0s 1ms/step - loss: 0.4020 - sparse_categorical_accuracy: 0.9268
Epoch 389/500
27/27 [==============================] - 0s 972us/step - loss: 0.4067 - sparse_categorical_accuracy: 0.9174
Epoch 390/500
27/27 [==============================] - 0s 947us/step - loss: 0.3943 - sparse_categorical_accuracy: 0.9303
Epoch 391/500
27/27 [==============================] - 0s 900us/step - loss: 0.3964 - sparse_categorical_accuracy: 0.9315
Epoch 392/500
27/27 [==============================] - 0s 1ms/step - loss: 0.3911 - sparse_categorical_accuracy: 0.9374
Epoch 393/500
27/27 [==============================] - 0s 1ms/step - loss: 0.3944 - sparse_categorical_accuracy: 0.9327
Epoch 394/500
27/27 [==============================] - 0s 861us/step - loss: 0.3830 - sparse_categorical_accuracy: 0.9268
Epoch 395/500
27/27 [==============================] - 0s 861us/step - loss: 0.3900 - sparse_categorical_accuracy: 0.9268
Epoch 396/500
27/27 [==============================] - 0s 956us/step - loss: 0.3933 - sparse_categorical_accuracy: 0.9386
Epoch 397/500
27/27 [==============================] - 0s 861us/step - loss: 0.3882 - sparse_categorical_accuracy: 0.9197
Epoch 398/500
27/27 [==============================] - 0s 981us/step - loss: 0.3919 - sparse_categorical_accuracy: 0.9292
Epoch 399/500
27/27 [==============================] - 0s 956us/step - loss: 0.3855 - sparse_categorical_accuracy: 0.9445
Epoch 400/500
27/27 [==============================] - 0s 2ms/step - loss: 0.3949 - sparse_categorical_accuracy: 0.9339 - val_loss: 0.4553 - val_sparse_categorical_accuracy: 0.9009
Epoch 401/500
27/27 [==============================] - 0s 976us/step - loss: 0.3855 - sparse_categorical_accuracy: 0.9327
Epoch 402/500
27/27 [==============================] - 0s 962us/step - loss: 0.3859 - sparse_categorical_accuracy: 0.9327
Epoch 403/500
27/27 [==============================] - 0s 999us/step - loss: 0.3948 - sparse_categorical_accuracy: 0.9292
Epoch 404/500
27/27 [==============================] - 0s 988us/step - loss: 0.3889 - sparse_categorical_accuracy: 0.9339
Epoch 405/500
27/27 [==============================] - 0s 986us/step - loss: 0.3880 - sparse_categorical_accuracy: 0.9268
Epoch 406/500
27/27 [==============================] - 0s 850us/step - loss: 0.4030 - sparse_categorical_accuracy: 0.9303
Epoch 407/500
27/27 [==============================] - 0s 873us/step - loss: 0.4043 - sparse_categorical_accuracy: 0.9162
Epoch 408/500
27/27 [==============================] - 0s 868us/step - loss: 0.3899 - sparse_categorical_accuracy: 0.9292
Epoch 409/500
27/27 [==============================] - 0s 853us/step - loss: 0.3889 - sparse_categorical_accuracy: 0.9351
Epoch 410/500
27/27 [==============================] - 0s 944us/step - loss: 0.3900 - sparse_categorical_accuracy: 0.9315
Epoch 411/500
27/27 [==============================] - 0s 947us/step - loss: 0.3880 - sparse_categorical_accuracy: 0.9292
Epoch 412/500
27/27 [==============================] - 0s 902us/step - loss: 0.3998 - sparse_categorical_accuracy: 0.9233
Epoch 413/500
27/27 [==============================] - 0s 867us/step - loss: 0.4004 - sparse_categorical_accuracy: 0.9209
Epoch 414/500
27/27 [==============================] - 0s 1ms/step - loss: 0.3790 - sparse_categorical_accuracy: 0.9386
Epoch 415/500
27/27 [==============================] - 0s 846us/step - loss: 0.3935 - sparse_categorical_accuracy: 0.9339
Epoch 416/500
27/27 [==============================] - 0s 922us/step - loss: 0.3920 - sparse_categorical_accuracy: 0.9256
Epoch 417/500
27/27 [==============================] - 0s 908us/step - loss: 0.4027 - sparse_categorical_accuracy: 0.9221
Epoch 418/500
27/27 [==============================] - 0s 896us/step - loss: 0.3829 - sparse_categorical_accuracy: 0.9315
Epoch 419/500
27/27 [==============================] - 0s 871us/step - loss: 0.3883 - sparse_categorical_accuracy: 0.9303
Epoch 420/500
27/27 [==============================] - 0s 899us/step - loss: 0.3951 - sparse_categorical_accuracy: 0.9244
Epoch 421/500
27/27 [==============================] - 0s 954us/step - loss: 0.3848 - sparse_categorical_accuracy: 0.9244
Epoch 422/500
27/27 [==============================] - 0s 849us/step - loss: 0.3833 - sparse_categorical_accuracy: 0.9327
Epoch 423/500
27/27 [==============================] - 0s 970us/step - loss: 0.3908 - sparse_categorical_accuracy: 0.9327
Epoch 424/500
27/27 [==============================] - 0s 882us/step - loss: 0.3799 - sparse_categorical_accuracy: 0.9398
Epoch 425/500
27/27 [==============================] - 0s 925us/step - loss: 0.3862 - sparse_categorical_accuracy: 0.9280
Epoch 426/500
27/27 [==============================] - 0s 908us/step - loss: 0.3862 - sparse_categorical_accuracy: 0.9362
Epoch 427/500
27/27 [==============================] - 0s 905us/step - loss: 0.3902 - sparse_categorical_accuracy: 0.9362
Epoch 428/500
27/27 [==============================] - 0s 943us/step - loss: 0.3926 - sparse_categorical_accuracy: 0.9327
Epoch 429/500
27/27 [==============================] - 0s 858us/step - loss: 0.3928 - sparse_categorical_accuracy: 0.9292
Epoch 430/500
27/27 [==============================] - 0s 923us/step - loss: 0.3794 - sparse_categorical_accuracy: 0.9362
Epoch 431/500
27/27 [==============================] - 0s 909us/step - loss: 0.4107 - sparse_categorical_accuracy: 0.9174
Epoch 432/500
27/27 [==============================] - 0s 921us/step - loss: 0.3922 - sparse_categorical_accuracy: 0.9303
Epoch 433/500
27/27 [==============================] - 0s 897us/step - loss: 0.3915 - sparse_categorical_accuracy: 0.9244
Epoch 434/500
27/27 [==============================] - 0s 890us/step - loss: 0.3866 - sparse_categorical_accuracy: 0.9327
Epoch 435/500
27/27 [==============================] - 0s 935us/step - loss: 0.3959 - sparse_categorical_accuracy: 0.9256
Epoch 436/500
27/27 [==============================] - 0s 881us/step - loss: 0.3969 - sparse_categorical_accuracy: 0.9268
Epoch 437/500
27/27 [==============================] - 0s 923us/step - loss: 0.4027 - sparse_categorical_accuracy: 0.9256
Epoch 438/500
27/27 [==============================] - 0s 936us/step - loss: 0.3999 - sparse_categorical_accuracy: 0.9315
Epoch 439/500
27/27 [==============================] - 0s 864us/step - loss: 0.3844 - sparse_categorical_accuracy: 0.9280
Epoch 440/500
27/27 [==============================] - 0s 894us/step - loss: 0.3873 - sparse_categorical_accuracy: 0.9268
Epoch 441/500
27/27 [==============================] - 0s 891us/step - loss: 0.3905 - sparse_categorical_accuracy: 0.9197
Epoch 442/500
27/27 [==============================] - 0s 903us/step - loss: 0.3849 - sparse_categorical_accuracy: 0.9386
Epoch 443/500
27/27 [==============================] - 0s 859us/step - loss: 0.3955 - sparse_categorical_accuracy: 0.9256
Epoch 444/500
27/27 [==============================] - 0s 879us/step - loss: 0.4151 - sparse_categorical_accuracy: 0.9091
Epoch 445/500
27/27 [==============================] - 0s 942us/step - loss: 0.4156 - sparse_categorical_accuracy: 0.9162
Epoch 446/500
27/27 [==============================] - 0s 928us/step - loss: 0.3875 - sparse_categorical_accuracy: 0.9315
Epoch 447/500
27/27 [==============================] - 0s 934us/step - loss: 0.4006 - sparse_categorical_accuracy: 0.9197
Epoch 448/500
27/27 [==============================] - 0s 928us/step - loss: 0.4075 - sparse_categorical_accuracy: 0.9126
Epoch 449/500
27/27 [==============================] - 0s 905us/step - loss: 0.3885 - sparse_categorical_accuracy: 0.9362
Epoch 450/500
27/27 [==============================] - 0s 937us/step - loss: 0.3912 - sparse_categorical_accuracy: 0.9233
Epoch 451/500
27/27 [==============================] - 0s 893us/step - loss: 0.3936 - sparse_categorical_accuracy: 0.9150
Epoch 452/500
27/27 [==============================] - 0s 909us/step - loss: 0.3911 - sparse_categorical_accuracy: 0.9303
Epoch 453/500
27/27 [==============================] - 0s 876us/step - loss: 0.3863 - sparse_categorical_accuracy: 0.9244
Epoch 454/500
27/27 [==============================] - 0s 910us/step - loss: 0.3830 - sparse_categorical_accuracy: 0.9268
Epoch 455/500
27/27 [==============================] - 0s 882us/step - loss: 0.3855 - sparse_categorical_accuracy: 0.9268
Epoch 456/500
27/27 [==============================] - 0s 896us/step - loss: 0.3845 - sparse_categorical_accuracy: 0.9280
Epoch 457/500
27/27 [==============================] - 0s 952us/step - loss: 0.3821 - sparse_categorical_accuracy: 0.9256
Epoch 458/500
27/27 [==============================] - 0s 891us/step - loss: 0.3823 - sparse_categorical_accuracy: 0.9292
Epoch 459/500
27/27 [==============================] - 0s 832us/step - loss: 0.3910 - sparse_categorical_accuracy: 0.9256
Epoch 460/500
27/27 [==============================] - 0s 905us/step - loss: 0.3842 - sparse_categorical_accuracy: 0.9315
Epoch 461/500
27/27 [==============================] - 0s 858us/step - loss: 0.3790 - sparse_categorical_accuracy: 0.9292
Epoch 462/500
27/27 [==============================] - 0s 856us/step - loss: 0.3917 - sparse_categorical_accuracy: 0.9162
Epoch 463/500
27/27 [==============================] - 0s 868us/step - loss: 0.3878 - sparse_categorical_accuracy: 0.9292
Epoch 464/500
27/27 [==============================] - 0s 901us/step - loss: 0.3862 - sparse_categorical_accuracy: 0.9339
Epoch 465/500
27/27 [==============================] - 0s 856us/step - loss: 0.3827 - sparse_categorical_accuracy: 0.9374
Epoch 466/500
27/27 [==============================] - 0s 931us/step - loss: 0.3956 - sparse_categorical_accuracy: 0.9221
Epoch 467/500
27/27 [==============================] - 0s 859us/step - loss: 0.3917 - sparse_categorical_accuracy: 0.9280
Epoch 468/500
27/27 [==============================] - 0s 967us/step - loss: 0.3924 - sparse_categorical_accuracy: 0.9221
Epoch 469/500
27/27 [==============================] - 0s 867us/step - loss: 0.3860 - sparse_categorical_accuracy: 0.9268
Epoch 470/500
27/27 [==============================] - 0s 981us/step - loss: 0.3791 - sparse_categorical_accuracy: 0.9315
Epoch 471/500
27/27 [==============================] - 0s 951us/step - loss: 0.4083 - sparse_categorical_accuracy: 0.9162
Epoch 472/500
27/27 [==============================] - 0s 864us/step - loss: 0.3800 - sparse_categorical_accuracy: 0.9268
Epoch 473/500
27/27 [==============================] - 0s 863us/step - loss: 0.4029 - sparse_categorical_accuracy: 0.9233
Epoch 474/500
27/27 [==============================] - 0s 837us/step - loss: 0.3787 - sparse_categorical_accuracy: 0.9433
Epoch 475/500
27/27 [==============================] - 0s 961us/step - loss: 0.3807 - sparse_categorical_accuracy: 0.9244
Epoch 476/500
27/27 [==============================] - 0s 901us/step - loss: 0.3782 - sparse_categorical_accuracy: 0.9351
Epoch 477/500
27/27 [==============================] - 0s 869us/step - loss: 0.3796 - sparse_categorical_accuracy: 0.9268
Epoch 478/500
27/27 [==============================] - 0s 909us/step - loss: 0.3827 - sparse_categorical_accuracy: 0.9374
Epoch 479/500
27/27 [==============================] - 0s 922us/step - loss: 0.3773 - sparse_categorical_accuracy: 0.9268
Epoch 480/500
27/27 [==============================] - 0s 880us/step - loss: 0.3920 - sparse_categorical_accuracy: 0.9292
Epoch 481/500
27/27 [==============================] - 0s 848us/step - loss: 0.3919 - sparse_categorical_accuracy: 0.9244
Epoch 482/500
27/27 [==============================] - 0s 858us/step - loss: 0.3953 - sparse_categorical_accuracy: 0.9162
Epoch 483/500
27/27 [==============================] - 0s 874us/step - loss: 0.3862 - sparse_categorical_accuracy: 0.9185
Epoch 484/500
27/27 [==============================] - 0s 862us/step - loss: 0.3797 - sparse_categorical_accuracy: 0.9339
Epoch 485/500
27/27 [==============================] - 0s 864us/step - loss: 0.4016 - sparse_categorical_accuracy: 0.9162
Epoch 486/500
27/27 [==============================] - 0s 946us/step - loss: 0.3896 - sparse_categorical_accuracy: 0.9339
Epoch 487/500
27/27 [==============================] - 0s 914us/step - loss: 0.3758 - sparse_categorical_accuracy: 0.9362
Epoch 488/500
27/27 [==============================] - 0s 891us/step - loss: 0.3861 - sparse_categorical_accuracy: 0.9303
Epoch 489/500
27/27 [==============================] - 0s 816us/step - loss: 0.3887 - sparse_categorical_accuracy: 0.9185
Epoch 490/500
27/27 [==============================] - 0s 809us/step - loss: 0.3923 - sparse_categorical_accuracy: 0.9233
Epoch 491/500
27/27 [==============================] - 0s 917us/step - loss: 0.3886 - sparse_categorical_accuracy: 0.9268
Epoch 492/500
27/27 [==============================] - 0s 905us/step - loss: 0.3853 - sparse_categorical_accuracy: 0.9221
Epoch 493/500
27/27 [==============================] - 0s 873us/step - loss: 0.3981 - sparse_categorical_accuracy: 0.9280
Epoch 494/500
27/27 [==============================] - 0s 946us/step - loss: 0.3809 - sparse_categorical_accuracy: 0.9315
Epoch 495/500
27/27 [==============================] - 0s 936us/step - loss: 0.3775 - sparse_categorical_accuracy: 0.9303
Epoch 496/500
27/27 [==============================] - 0s 908us/step - loss: 0.3800 - sparse_categorical_accuracy: 0.9386
Epoch 497/500
27/27 [==============================] - 0s 962us/step - loss: 0.3818 - sparse_categorical_accuracy: 0.9303
Epoch 498/500
27/27 [==============================] - 0s 939us/step - loss: 0.3777 - sparse_categorical_accuracy: 0.9256
Epoch 499/500
27/27 [==============================] - 0s 909us/step - loss: 0.3881 - sparse_categorical_accuracy: 0.9327
Epoch 500/500
27/27 [==============================] - 0s 2ms/step - loss: 0.3823 - sparse_categorical_accuracy: 0.9339 - val_loss: 0.4249 - val_sparse_categorical_accuracy: 0.9151

查看网络结构

# 查看网络结构
model.summary()

Model: "sequential_3"
_________________________________________________________________
 Layer (type)                Output Shape              Param #   
=================================================================
 dense_9 (Dense)             (None, 30)                240       
                                                                 
 dense_10 (Dense)            (None, 90)                2790      
                                                                 
 dense_11 (Dense)            (None, 3)                 273       
                                                                 
=================================================================
Total params: 3,303
Trainable params: 3,303
Non-trainable params: 0
_________________________________________________________________

进行模型评估

# 划分一些测试集出来，选择后300行作为测试集
features_test = features_data[-300:]
labels_test = labels_data[-300:]
# 输出测试集
# print(features_test)
# print(labels_test)

# 评估模型
loss, accuracy = model.evaluate(features_test, labels_test)
print(f'模型的准确率为：{accuracy}')

10/10 [==============================] - 0s 997us/step - loss: 0.4143 - sparse_categorical_accuracy: 0.9167
模型的准确率为：0.9166666865348816

总结与分析

本文通过TensorFlow框架中的keras构建简单神经网络对牛奶质量进行分类（预测），首先对实验数据集进行预处理，将对象类型的数据转换为数值类型，并提取出特征列和标签列，以便于后续模型的训练；网络结构为三层全连接层， 第一层有30个神经元，激活函数为relu，第二层有90个神经元，激活函数为relu，第三层，也是输出层，有3个神经元，激活函数为softmax，更多的详细参数配置在上文分类过程中有所体现；通过对模型的评估，得到了超过90%的准确率，有一定的可用性，后续还可以通过调整模型参数提升效果。

你可能感兴趣的:(tensorflow,keras,神经网络,分类,深度学习)

  • 机器学习与深度学习间关系与区别 ℒℴѵℯ心·动ꦿ໊ོ꫞ 人工智能学习深度学习python
    一、机器学习概述定义机器学习（MachineLearning,ML）是一种通过数据驱动的方法，利用统计学和计算算法来训练模型，使计算机能够从数据中学习并自动进行预测或决策。机器学习通过分析大量数据样本，识别其中的模式和规律，从而对新的数据进行判断。其核心在于通过训练过程，让模型不断优化和提升其预测准确性。主要类型1.监督学习（SupervisedLearning）监督学习是指在训练数据集中包含输入
  • 将cmd中命令输出保存为txt文本文件 落难Coder Windowscmdwindow
    最近深度学习本地的训练中我们常常要在命令行中运行自己的代码，无可厚非，我们有必要保存我们的炼丹结果，但是复制命令行输出到txt是非常麻烦的，其实Windows下的命令行为我们提供了相应的操作。其基本的调用格式就是：运行指令>输出到的文件名称或者具体保存路径测试下，我打开cmd并且ping一下百度：pingwww.baidu.com>./data.txt看下相同目录下data.txt的输出：如果你再
  • 番茄西红柿叶子病害分类数据集12882张11类别 futureflsl 数据集分类数据挖掘人工智能
    数据集类型：图像分类用，不可用于目标检测无标注文件数据集格式：仅仅包含jpg图片，每个类别文件夹下面存放着对应图片图片数量(jpg文件个数)：12882分类类别数：11类别名称:["Bacterial_Spot_Bacteria","Early_Blight_Fungus","Healthy","Late_Blight_Water_Mold","Leaf_Mold_Fungus","Powdery
  • Python开发常用的三方模块如下： 换个网名有点难 python开发语言
    Python是一门功能强大的编程语言，拥有丰富的第三方库，这些库为开发者提供了极大的便利。以下是100个常用的Python库，涵盖了多个领域：1、NumPy，用于科学计算的基础库。2、Pandas，提供数据结构和数据分析工具。3、Matplotlib，一个绘图库。4、Scikit-learn，机器学习库。5、SciPy，用于数学、科学和工程的库。6、TensorFlow，由Google开发的开源机
  • xilinx vivado PULLMODE 设置思路 坚持每天写程序 fpga开发
    1.xilinx引脚分类XilinxIO的分类：以XC7A100TFGG484为例，其引脚分类如下：1.UserIO(用户IO)：用户使用的普通IO1.1专用(Dedicated)IO：命名为IO_LXXY_#、IO_XX_#的引脚，有固定的特定用途，多为底层特定功能的直接实现，如差分对信号、关键控制信号等，不能随意变更。1.2多功能(Multi-Function)IO：命名为IO_LXXY_ZZ
  • Python实现简单的机器学习算法 master_chenchengg pythonpython办公效率python开发IT
    Python实现简单的机器学习算法开篇：初探机器学习的奇妙之旅搭建环境：一切从安装开始必备工具箱第一步：安装Anaconda和JupyterNotebook小贴士：如何配置Python环境变量算法初体验：从零开始的Python机器学习线性回归：让数据说话数据准备：从哪里找数据编码实战：Python实现线性回归模型评估：如何判断模型好坏逻辑回归：从分类开始理论入门：什么是逻辑回归代码实现：使用skl
  • K近邻算法_分类鸢尾花数据集 _feivirus_ 算法机器学习和数学分类机器学习K近邻
    importnumpyasnpimportpandasaspdfromsklearn.datasetsimportload_irisfromsklearn.model_selectionimporttrain_test_splitfromsklearn.metricsimportaccuracy_score1.数据预处理iris=load_iris()df=pd.DataFrame(data=ir
  • 网络通信流程 记得开心一点啊 服务器网络运维
    目录♫IP地址♫子网掩码♫MAC地址♫相关设备♫ARP寻址♫网络通信流程♫IP地址我们已经知道IP地址由网络号+主机号组成，根据IP地址的不同可以有5钟划分网络号和主机号的方案：其中，各类地址的表示范围是：分类范围适用网络网络数量主机最大连接数A类0.0.0.0~127.255.255.255大型网络12616777214【(2^24)-2】B类128.0.0.0~191.255.255.255中
  • 5分钟说透AppStore审核原理，让你拥有上架新思路！ Q仔本人噢
    在AppStore上架是越来越难了!相信非常多公司的技术人员都为此困扰，然而外包团队水平又层次不齐，容易遇坑，实在是内忧外患。是什么原因导致审核机制频繁调整？又是什么原因使得审核变得越发严格？那么接下来听小Q分解，马上给各位带来解答!首先看一下近一年的上下架的情况：近一年上架情况近一年下架情况通过数据我们发现越是马甲包产量权重高的分类里被下架的app数量越多，苹果此举可谓是上有政策，下有对策。通过
  • 遥感影像的切片处理 sand&wich 计算机视觉python图像处理
    在遥感影像分析中，经常需要将大尺寸的影像切分成小片段，以便于进行详细的分析和处理。这种方法特别适用于机器学习和图像处理任务，如对象检测、图像分类等。以下是如何使用Python和OpenCV库来实现这一过程，同时确保每个影像片段保留正确的地理信息。准备环境首先，确保安装了必要的Python库，包括numpy、opencv-python和xml.etree.ElementTree。这些库将用于图像处理
  • 《转介绍方法论》学习笔记 小可乐的妈妈
    一、高效转介绍的流程：价值观---执行----方案一）转介绍发生的背景：1、对象：谁向谁转介绍？全员营销，人人参与。①员工的激励政策、客户的转介绍诱因制作客户画像：a信任；支付能力；意愿度；便利度（根据家长具备四个特征的个数分为四类）B性格分类C职业分类D年龄性别②执行：套路，策略，方法，流程2、诱因：为什么要转介绍？认同信任；多方共赢；传递美好；零风险承诺打动人心，超越期待。选择做教育，就是选择
  • JAVA学习笔记之23种设计模式学习 victorfreedom Java技术设计模式androidjava常用设计模式
    博主最近买了《设计模式》这本书来学习，无奈这本书是以C++语言为基础进行说明，整个学习流程下来效率不是很高，虽然有的设计模式通俗易懂，但感觉还是没有充分的掌握了所有的设计模式。于是博主百度了一番，发现有大神写过了这方面的问题，于是博主迅速拿来学习。一、设计模式的分类总体来说设计模式分为三大类：创建型模式，共五种：工厂方法模式、抽象工厂模式、单例模式、建造者模式、原型模式。结构型模式，共七种：适配器
  • 100天持续行动—Day01 Richard_DL
    今天开始站着学习，发现效率大幅提升。把fast.ai的Lesson1的后半部分和Lesson2看完了。由于Keras版本和视频中的不一致，运行notebook时经常出现莫名其妙的错误，导致自己只动手实践了视频中的一小部分内容。为了赶时间，我打算先把与CNN相关的视频过一遍。然后尽快开始做自己的项目。明天继续加油，争取把Lesson3和Lesson4看完。
  • 推荐3家毕业AI论文可五分钟一键生成！文末附免费教程！ 小猪包333 写论文人工智能AI写作深度学习计算机视觉
    在当前的学术研究和写作领域，AI论文生成器已经成为许多研究人员和学生的重要工具。这些工具不仅能够帮助用户快速生成高质量的论文内容，还能进行内容优化、查重和排版等操作。以下是三款值得推荐的AI论文生成器：千笔-AIPassPaper、懒人论文以及AIPaperPass。千笔-AIPassPaper千笔-AIPassPaper是一款基于深度学习和自然语言处理技术的AI写作助手，旨在帮助用户快速生成高质
  • 【自动化测试】UI自动化的分类、如何选择合适的自动化测试工具以及其中appium的设计理念、引擎和引擎如何工作 Lossya ui自动化测试工具自动化测试appium
    引言UI自动化测试主要针对软件的用户界面进行测试，以确保用户界面元素的交互和功能符合预期文章目录引言一、UI自动化的分类1.1基于代码的自动化测试1.2基于录制/回放的自动化测试1.3基于框架的自动化测试1.4按测试对象分类1.5按测试层次分类1.6按测试执行方式分类1.7按测试目的分类二、如何选择合适的自动化测试工具2.1项目需求分析2.2工具特性评估2.3成本考虑2.4团队技能2.5试用和评估
  • AI大模型的架构演进与最新发展 季风泯灭的季节 AI大模型应用技术二人工智能架构
    随着深度学习的发展，AI大模型（LargeLanguageModels,LLMs）在自然语言处理、计算机视觉等领域取得了革命性的进展。本文将详细探讨AI大模型的架构演进，包括从Transformer的提出到GPT、BERT、T5等模型的历史演变，并探讨这些模型的技术细节及其在现代人工智能中的核心作用。一、基础模型介绍：Transformer的核心原理Transformer架构的背景在Transfo
  • ai绘画工具midjourney怎么下载？附作品管理教程 设计师早上好
    Midjourney是一款功能强大的AI绘画工具，它使用机器学习技术和深度神经网络等算法，可以生成各种艺术风格的绘画作品。在创意设计、广告宣传等方面有着广泛的应用前景。那么，ai绘画工具midjourney怎么下载？本文将为您介绍Midjourney的下载以及作品的相关管理。一、Midjourney下载Midjourney的下载非常简单，只需打开Midjourney官网（点击“GetMidjour
  • 性格小测试 熹大头
    有些人非常肯定自己属于外向型，有些人则发现自己是绝对的内向型。然而，多数人却发现他们似乎介于两者之间，是两种性格的结合。现在我们就来看看你在这种分类中处在何种位置。阅读以下问题，从a、b、c中选出最适合自己的选项。你可能会发现三个选项都不合适，或者合适的不止一项，这种情况下，选出相对来说更适合自己的即可。1人们经常会用下列哪个词语描述你：a善于分析b遵守纪律c有创造力2一连几天参与社交活动（比如，
  • 李克富 | 咨询师推荐阅读书目 李克富
    最重要的书籍不是别人的推荐，而是自己学过的教材，不论当初使用的是哪个版本，它都是我们专业的底层代码，具有不可替代性。前不久，中国心理咨询师筹委会的一位老师邀请我罗列一个推荐书目清单作为咨询师工具包的内容，并要求“说明一下简单的分类或者作三言两语的说明”。斟酌后，我觉得自己推荐的书目大体可以分为普及类书籍、心理学书籍和心理咨询与治疗专业书籍，第三类又分为适合于咨询师新手的和有经验咨询师的。经过严格筛
  • [实践应用] 深度学习之模型性能评估指标 YuanDaima2048 深度学习工具使用深度学习人工智能损失函数性能评估pytorchpython机器学习
    文章总览：YuanDaiMa2048博客文章总览深度学习之模型性能评估指标分类任务回归任务排序任务聚类任务生成任务其他介绍在机器学习和深度学习领域，评估模型性能是一项至关重要的任务。不同的学习任务需要不同的性能指标来衡量模型的有效性。以下是对一些常见任务及其相应的性能评估指标的详细解释和总结。分类任务分类任务是指模型需要将输入数据分配到预定义的类别或标签中。以下是分类任务中常用的性能指标：准确率(
  • [实践应用] 深度学习之优化器 YuanDaima2048 深度学习工具使用pytorch深度学习人工智能机器学习python优化器
    文章总览：YuanDaiMa2048博客文章总览深度学习之优化器1.随机梯度下降（SGD）2.动量优化（Momentum）3.自适应梯度（Adagrad）4.自适应矩估计（Adam）5.RMSprop总结其他介绍在深度学习中，优化器用于更新模型的参数，以最小化损失函数。常见的优化函数有很多种，下面是几种主流的优化器及其特点、原理和PyTorch实现：1.随机梯度下降（SGD）原理:随机梯度下降通过
  • 新能源汽车 BMS 学习笔记篇—BMS 基本定义及分类 WPG大大通 其他笔记汽车BMS经验分享新能源电池
    一、BMS定义1、概念：BMS（BatteryManagementSystem）即电池管理系统，其管理对象是二次电池（充电电池或蓄电池），其主要目的是电池的利用率，防止电池出现过度充电和过度放电，可应用于电动汽车、电瓶车、机器人、无人机等图片来源：腾讯网https://new.qq.com《标准普尔警告，电动汽车电池生产面临供应链和地缘政治风险》2、四大功能①感知和测量：检测电池的电压、电流、温度
  • 生成式地图制图 Bwywb_3 深度学习机器学习深度学习生成对抗网络
    生成式地图制图（GenerativeCartography）是一种利用生成式算法和人工智能技术自动创建地图的技术。它结合了传统的地理信息系统（GIS）技术与现代生成模型（如深度学习、GANs等），能够根据输入的数据自动生成符合需求的地图。这种方法在城市规划、虚拟环境设计、游戏开发等多个领域具有应用前景。主要特点：自动化生成：通过算法和模型，系统能够根据输入的地理或空间数据自动生成地图，而无需人工逐
  • 郭生白中药方论之二(破除温凉寒热的框框) 本能学堂a昨年
    离病说药茫茫然，对症下药不着边。顺势利导一乘法，排异调节渡法船。无限整合非模糊，模糊病区得清楚。共性之外求个性，亲和不生抗药性。温凉寒热巧方便，君臣佐使筏喻焉。药包大小折中看，毒性有无一念间。导读破除温凉寒热的框框寒热温凉是基于中药共性的传统分类药无寒热人有寒热药无寒热病有寒热抛弃温凉不并用的错误观念寒热温凉是基于中药共性的传统分类寒热温凉是个共性，是说的共性。这个共性，知道什么叫共性吗？所有的药
  • 2022-04-25 L是木子李呢
    上门维修APP开发应具备哪些功能随着移动互联网的不断发展，上门维修在我们生活中已经是非常普遍的存在了，为了给用户更方便的找到上门维修的渠道，上门维修APP应运而生，那么上门维修APP开发应具备哪些功能呢？1、维修门店搜索为了更好地方便用户省时省力，上门维修APP会依据用户定位信息搜索线下实体店，促使用户更好的找到线下维修店面，省时又省力。2、维修服务分类包括管道洁具维修、强电弱电维修、木工维修、粉
  • 吴恩达深度学习笔记(30)-正则化的解释 极客Array
    正则化（Regularization）深度学习可能存在过拟合问题——高方差，有两个解决方法，一个是正则化，另一个是准备更多的数据，这是非常可靠的方法，但你可能无法时时刻刻准备足够多的训练数据或者获取更多数据的成本很高，但正则化通常有助于避免过拟合或减少你的网络误差。如果你怀疑神经网络过度拟合了数据，即存在高方差问题，那么最先想到的方法可能是正则化，另一个解决高方差的方法就是准备更多数据，这也是非常
  • 个人学习笔记7-6：动手学深度学习pytorch版-李沐 浪子L 深度学习深度学习笔记计算机视觉python人工智能神经网络pytorch
    #人工智能##深度学习##语义分割##计算机视觉##神经网络#计算机视觉13.11全卷积网络全卷积网络（fullyconvolutionalnetwork，FCN）采用卷积神经网络实现了从图像像素到像素类别的变换。引入l转置卷积（transposedconvolution）实现的，输出的类别预测与输入图像在像素级别上具有一一对应关系：通道维的输出即该位置对应像素的类别预测。13.11.1构造模型下
  • 二十四、k8s 资源管理 繁华依在 k8skubernetes容器云原生
    目录一、资源配置范围管理LimitRange介绍1、LimitRange可以做什么：2、资源限制和请求的约束3、创建LimitsRange对象4、示例：创建一个pod5、测试用例测试1：测试2：测试3：二、资源服务质量管理（RequestsQos）1、Qos级别分类：1.1、Guaranteed：1.2、BestEffort：1.3、Burstable：2、Qos的工作特点3、示例三、资源配额管理
  • FlagEmbedding 吉小雨 python库python
    FlagEmbedding教程FlagEmbedding是一个用于生成文本嵌入（textembeddings）的库，适合处理自然语言处理（NLP）中的各种任务。嵌入（embeddings）是将文本表示为连续向量，能够捕捉语义上的相似性，常用于文本分类、聚类、信息检索等场景。官方文档链接：FlagEmbedding官方GitHub一、FlagEmbedding库概述1.1什么是FlagEmbeddi
  • 深度学习-点击率预估-研究论文2024-09-14速读 sp_fyf_2024 深度学习人工智能
    深度学习-点击率预估-研究论文2024-09-14速读1.DeepTargetSessionInterestNetworkforClick-ThroughRatePredictionHZhong,JMa,XDuan,SGu,JYao-2024InternationalJointConferenceonNeuralNetworks,2024深度目标会话兴趣网络用于点击率预测摘要：这篇文章提出了一种新
  • 数据采集高并发的架构应用 3golden .net
    问题的出发点：          最近公司为了发展需要，要扩大对用户的信息采集，每个用户的采集量估计约2W。如果用户量增加的话，将会大量照成采集量成3W倍的增长，但是又要满足日常业务需要，特别是指令要及时得到响应的频率次数远大于预期。       &n
  • 不停止 MySQL 服务增加从库的两种方式 brotherlamp linuxlinux视频linux资料linux教程linux自学
    现在生产环境MySQL数据库是一主一从，由于业务量访问不断增大，故再增加一台从库。前提是不能影响线上业务使用，也就是说不能重启MySQL服务，为了避免出现其他情况，选择在网站访问量低峰期时间段操作。  一般在线增加从库有两种方式，一种是通过mysqldump备份主库，恢复到从库，mysqldump是逻辑备份，数据量大时，备份速度会很慢，锁表的时间也会很长。另一种是通过xtrabacku
  • Quartz——SimpleTrigger触发器 eksliang SimpleTriggerTriggerUtilsquartz
    转载请出自出处：http://eksliang.iteye.com/blog/2208166 一.概述 SimpleTrigger触发器，当且仅需触发一次或者以固定时间间隔周期触发执行；   二.SimpleTrigger的构造函数 SimpleTrigger(String name, String group)：通过该构造函数指定Trigger所属组和名称； Simpl
  • Informatica应用（1） 18289753290 sqlworkflowlookup组件Informatica
    1.如果要在workflow中调用shell脚本有一个command组件，在里面设置shell的路径；调度wf可以右键出现schedule，现在用的是HP的tidal调度wf的执行。 2.designer里面的router类似于SSIS中的broadcast（多播组件）;Reset_Workflow_Var：参数重置 （比如说我这个参数初始是1在workflow跑得过程中变成了3我要在结束时还要
  • python 获取图片验证码中文字 酷的飞上天空 python
    根据现成的开源项目 http://code.google.com/p/pytesser/改写 在window上用easy_install安装不上  看了下源码发现代码很少  于是就想自己改写一下   添加支持网络图片的直接解析   #coding:utf-8 #import sys #reload(sys) #sys.s
  • AJAX 永夜-极光 Ajax
    1.AJAX功能:动态更新页面,减少流量消耗,减轻服务器负担   2.代码结构:    <html> <head> <script type="text/javascript"> function loadXMLDoc() { .... AJAX script goes here ...
  • 创业OR读研 随便小屋 创业
            现在研一，有种想创业的想法，不知道该不该去实施。因为对于的我情况这两者是矛盾的，可能就是鱼与熊掌不能兼得。                研一的生活刚刚过去两个月，我们学校主要的是
  • 需求做得好与坏直接关系着程序员生活质量 aijuans IT 生活
            这个故事还得从去年换工作的事情说起，由于自己不太喜欢第一家公司的环境我选择了换一份工作。去年九月份我入职现在的这家公司，专门从事金融业内软件的开发。十一月份我们整个项目组前往北京做现场开发，从此苦逼的日子开始了。        系统背景：五月份就有同事前往甲方了解需求一直到6月份，后续几个月也完
  • 如何定义和区分高级软件开发工程师 aoyouzi
    在软件开发领域，高级开发工程师通常是指那些编写代码超过 3 年的人。这些人可能会被放到领导的位置，但经常会产生非常糟糕的结果。Matt Briggs 是一名高级开发工程师兼 Scrum 管理员。他认为，单纯使用年限来划分开发人员存在问题，两个同样具有 10 年开发经验的开发人员可能大不相同。近日，他发表了一篇博文，根据开发者所能发挥的作用划分软件开发工程师的成长阶段。 　　初
  • Servlet的请求与响应 百合不是茶 servletget提交java处理post提交
      Servlet是tomcat中的一个重要组成,也是负责客户端和服务端的中介     1,Http的请求方式(get  ,post);   客户端的请求一般都会都是Servlet来接受的,在接收之前怎么来确定是那种方式提交的,以及如何反馈,Servlet中有相应的方法,  http的get方式 servlet就是都doGet(
  • web.xml配置详解之listener bijian1013 javaweb.xmllistener
    一.定义 <listener> <listen-class>com.myapp.MyListener</listen-class> </listener>   二.作用        该元素用来注册一个监听器类。可以收到事件什么时候发生以及用什么作为响
  • Web页面性能优化（yahoo技术） Bill_chen JavaScriptAjaxWebcssYahoo
    1.尽可能的减少HTTP请求数 content 2.使用CDN server 3.添加Expires头(或者 Cache-control) server 4.Gzip 组件 server 5.把CSS样式放在页面的上方。 css 6.将脚本放在底部(包括内联的) javascript 7.避免在CSS中使用Expressions css 8.将javascript和css独立成外部文
  • 【MongoDB学习笔记八】MongoDB游标、分页查询、查询结果排序 bit1129 mongodb
    游标   游标，简单的说就是一个查询结果的指针。游标作为数据库的一个对象，使用它是包括 声明 打开 循环抓去一定数目的文档直到结果集中的所有文档已经抓取完 关闭游标   游标的基本用法，类似于JDBC的ResultSet(hasNext判断是否抓去完,next移动游标到下一条文档)，在获取一个文档集时，可以提供一个类似JDBC的FetchSize
  • ORA-12514 TNS 监听程序当前无法识别连接描述符中请求服务 的解决方法 白糖_ ORA-12514
    今天通过Oracle SQL*Plus连接远端服务器的时候提示“监听程序当前无法识别连接描述符中请求服务”，遂在网上找到了解决方案： ①打开Oracle服务器安装目录\NETWORK\ADMIN\listener.ora文件，你会看到如下信息：   # listener.ora Network Configuration File: D:\database\Oracle\net
  • Eclipse 问题 A resource exists with a different case bozch eclipse
    在使用Eclipse进行开发的时候，出现了如下的问题： Description Resource Path Location TypeThe project was not built due to "A resource exists with a different case: '/SeenTaoImp_zhV2/bin/seentao'.&
  • 编程之美-小飞的电梯调度算法 bylijinnan 编程之美
    public class AptElevator { /** * 编程之美 小飞 电梯调度算法 * 在繁忙的时间，每次电梯从一层往上走时，我们只允许电梯停在其中的某一层。 * 所有乘客都从一楼上电梯，到达某层楼后，电梯听下来，所有乘客再从这里爬楼梯到自己的目的层。 * 在一楼时，每个乘客选择自己的目的层，电梯则自动计算出应停的楼层。 * 问：电梯停在哪
  • SQL注入相关概念 chenbowen00 sqlWeb安全
    SQL Injection：就是通过把SQL命令插入到Web表单递交或输入域名或页面请求的查询字符串，最终达到欺骗服务器执行恶意的SQL命令。 具体来说，它是利用现有应用程序，将（恶意）的SQL命令注入到后台数据库引擎执行的能力，它可以通过在Web表单中输入（恶意）SQL语句得到一个存在安全漏洞的网站上的数据库，而不是按照设计者意图去执行SQL语句。 首先让我们了解什么时候可能发生SQ
  • [光与电]光子信号战防御原理 comsci 原理
          无论是在战场上,还是在后方,敌人都有可能用光子信号对人体进行控制和攻击,那么采取什么样的防御方法,最简单,最有效呢?       我们这里有几个山寨的办法,可能有些作用,大家如果有兴趣可以去实验一下       根据光
  • oracle 11g新特性:Pending Statistics daizj oracledbms_stats
    oracle 11g新特性:Pending Statistics 转 从11g开始，表与索引的统计信息收集完毕后，可以选择收集的统信息立即发布，也可以选择使新收集的统计信息处于pending状态，待确定处于pending状态的统计信息是安全的，再使处于pending状态的统计信息发布，这样就会避免一些因为收集统计信息立即发布而导致SQL执行计划走错的灾难。 在 11g 之前的版本中，D
  • 快速理解RequireJs dengkane jqueryrequirejs
    RequireJs已经流行很久了，我们在项目中也打算使用它。它提供了以下功能： 声明不同js文件之间的依赖 可以按需、并行、延时载入js库 可以让我们的代码以模块化的方式组织 初看起来并不复杂。 在html中引入requirejs 在HTML中，添加这样的 <script> 标签： <script src="/path/to
  • C语言学习四流程控制if条件选择、for循环和强制类型转换 dcj3sjt126com c
    # include <stdio.h> int main(void) { int i, j; scanf("%d %d", &i, &j); if (i > j) printf("i大于j\n"); else printf("i小于j\n"); retu
  • dictionary的使用要注意 dcj3sjt126com IO
    NSDictionary *dict = [NSDictionary dictionaryWithObjectsAndKeys: user.user_id , @"id", user.username , @"username",
  • Android 中的资源访问(Resource) finally_m xmlandroidStringdrawablecolor
    简单的说，Android中的资源是指非代码部分。例如，在我们的Android程序中要使用一些图片来设置界面，要使用一些音频文件来设置铃声，要使用一些动画来显示特效，要使用一些字符串来显示提示信息。那么，这些图片、音频、动画和字符串等叫做Android中的资源文件。 在Eclipse创建的工程中，我们可以看到res和assets两个文件夹，是用来保存资源文件的，在assets中保存的一般是原生
  • Spring使用Cache、整合Ehcache 234390216 springcacheehcache@Cacheable
    Spring使用Cache            从3.1开始，Spring引入了对Cache的支持。其使用方法和原理都类似于Spring对事务管理的支持。Spring Cache是作用在方法上的，其核心思想是这样的：当我们在调用一个缓存方法时会把该方法参数和返回结果作为一个键值对存放在缓存中，等到下次利用同样的
  • 当druid遇上oracle blob(clob) jackyrong oracle
    http://blog.csdn.net/renfufei/article/details/44887371 众所周知，Oracle有很多坑, 所以才有了去IOE。 在使用Druid做数据库连接池后，其实偶尔也会碰到小坑，这就是使用开源项目所必须去填平的。【如果使用不开源的产品，那就不是坑，而是陷阱了，你都不知道怎么去填坑】 用Druid连接池，通过JDBC往Oracle数据库的
  • easyui datagrid pagination获得分页页码、总页数等信息 ldzyz007
    var grid = $('#datagrid');  var options = grid.datagrid('getPager').data("pagination").options;  var curr = options.pageNumber;  var total = options.total;  var max =
  • 浅析awk里的数组 nigelzeng 二维数组array数组awk
    awk绝对是文本处理中的神器，它本身也是一门编程语言，还有许多功能本人没有使用到。这篇文章就单单针对awk里的数组来进行讨论，如何利用数组来帮助完成文本分析。   有这么一组数据：   abcd,91#31#2012-12-31 11:24:00 case_a,136#19#2012-12-31 11:24:00 case_a,136#23#2012-12-31 1
  • 搭建 CentOS 6 服务器(6) - TigerVNC rensanning centos
    安装GNOME桌面环境 # yum groupinstall "X Window System" "Desktop" 安装TigerVNC # yum -y install tigervnc-server tigervnc 启动VNC服务 # /etc/init.d/vncserver restart # vncser
  • Spring 数据库连接整理 tomcat_oracle springbeanjdbc
    1、数据库连接jdbc.properties配置详解 　　jdbc.url=jdbc:hsqldb:hsql://localhost/xdb 　　jdbc.username=sa 　　jdbc.password= 　　jdbc.driver=不同的数据库厂商驱动，此处不一一列举 　　接下来，详细配置代码如下： 　　 Spring连接池    
  • Dom4J解析使用xpath java.lang.NoClassDefFoundError: org/jaxen/JaxenException异常 xp9802
    用Dom4J解析xml,以前没注意,今天使用dom4j包解析xml时在xpath使用处报错      异常栈：java.lang.NoClassDefFoundError: org/jaxen/JaxenException异常       导入包 jaxen-1.1-beta-6.jar 解决; &nb
按字母分类： ABCDEFGHIJKLMNOPQRSTUVWXYZ其他