【实验练习】基于SVM的实现鸢尾花(Iris)数据集分类 (Python实现)
题目
采用SVM方法实现鸢尾花(Iris)数据集分类
数据集
iris.name是关于数据集的属性说明；
iris.data是实际的数据集，它包含三类数据，每类数据有50条数据。
要求
训练集：选取Iris数据集中80%的数据，即120个数据，每类含有40个数据。
测试集：采用除训练集外的30个数据。
具体SVM方法：自由根据情况来选择。
评价指标：选取分类相关的评价指标来衡量分类结果。
目录结构
|----Exp3\
| |----Data\
| | |----iris.data
| | |----iris.names
| |----main.py
| |----Plot.py
| |----Result\
| | |----iris-cla.png
| | |----iris-petal-cla.png
| | |----iris-sepal-cal.png
| | |----iris-sepal-petal-cla.png
main.py
# -*- coding: utf-8 -*- #
"""
@Project ：MachineLearningLesson
@File ：main.py
@Author ：ZAY
@Time ：2023/6/4 15:44
@Annotation : " "
"""
import os
import torch
import sklearn
import numpy as np
from sklearn import svm
from sklearn.metrics import accuracy_score,auc,roc_curve,precision_recall_curve,f1_score, precision_score, recall_score
# Exp.Exp3为本项目存储路径，根据自己实际存储地址进行更改
from Exp.Exp3.Plot import plotSepalShow,plotPetalShow,plotSPShow
os.environ["CUDA_VISIBLE_DEVICES"] = "0"
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
# 将标签由文字映射为数字
def Iris_label(s):
it = {b'Iris-setosa': 0, b'Iris-versicolor': 1, b'Iris-virginica': 2}
return it[s]
def Iris_Sepal_Cla(data_x, label_y):
data_x = data_x[:, 0:2]
train_data, test_data, train_label, test_label = sklearn.model_selection.train_test_split(data_x, label_y,
random_state = 1,
train_size = 0.8,
test_size = 0.2)
# 训练 SVM 分类器 https://blog.csdn.net/TeFuirnever/article/details/99646257
classifier = svm.SVC(C = 0.5, kernel = 'rbf', gamma = 10, decision_function_shape = 'ovr') # rbf
classifier.fit(train_data, train_label.ravel())
train_label_pre = classifier.predict(train_data)
test_label_pre = classifier.predict(test_data)
print('花萼训练集acc:', accuracy_score(train_label, train_label_pre))
print('花萼测试集acc:', accuracy_score(test_label, test_label_pre))
# 查看内部决策函数（返回的是样本到超平面的距离）
train_decision_function = classifier.decision_function(train_data)
predict_result = classifier.predict(train_data)
# print('train_decision_function:', train_decision_function)
# print('predict_result:', predict_result)
plotSepalShow(test_data, test_label, data_x, label_y, classifier)
def Iris_Petal_Cla(data_x, label_y):
# 基于SVM鸢尾花瓣长宽度二特征分类
data_x = data_x[:, 2:4]
train_data, test_data, train_label, test_label = sklearn.model_selection.train_test_split(data_x, label_y,
random_state = 1,
train_size = 0.8,
test_size = 0.2)
classifier = svm.SVC(C = 0.5, kernel = 'rbf', gamma = 10, decision_function_shape = 'ovr') # rbf
classifier.fit(train_data, train_label.ravel())
train_label_pre = classifier.predict(train_data)
test_label_pre = classifier.predict(test_data)
print('花瓣训练集acc:', accuracy_score(train_label, train_label_pre))
print('花瓣测试集acc:', accuracy_score(test_label, test_label_pre))
plotPetalShow(test_data, test_label, data_x, label_y, classifier)
if __name__ == "__main__":
txt_path = './/Result//SVM.txt'
data = np.loadtxt("./Data/iris.data", dtype = float, delimiter = ',', converters = {4: Iris_label})
# 基于SVM鸢尾花萼长宽度二特征分类
data_x, label_y = np.split(data, indices_or_sections = (4,), axis = 1) # x为数据，y为标签
Iris_Sepal_Cla(data_x, label_y)
Iris_Petal_Cla(data_x, label_y)
plotSepalShow.py plotPetalShow.py
# -*- coding: utf-8 -*- #
"""
@Project ：MachineLearningLesson
@File ：plot.py
@Author ：ZAY
@Time ：2023/6/5 21:41
@Annotation : " "
"""
# 确定坐标轴范围
import matplotlib
import matplotlib.pyplot as plt
import numpy as np
def plotSepalShow(test_data, test_label, data_x, label_y, classifier):
x1_min, x1_max = data_x[:, 0].min(), data_x[:, 0].max() # 第0维特征的范围
x2_min, x2_max = data_x[:, 1].min(), data_x[:, 1].max() # 第1维特征的范围
x1, x2 = np.mgrid[x1_min:x1_max:200j, x2_min:x2_max:200j] # 生成网络采样点
# print(x1.shape) # (200, 200)
# print(x1.flat) # flat属性可以使得像遍历以一维数组的方法来遍历多维数组
grid_test = np.stack((x1.flat, x2.flat), axis = 1) # 测试点
# print(grid_test.shape) # (40000, 2)
# 指定默认字体
matplotlib.rcParams['font.sans-serif'] = ['SimHei']
# 设置颜色
cm_light = matplotlib.colors.ListedColormap(['#A0FFA0', '#FFA0A0', '#A0A0FF'])
cm_dark = matplotlib.colors.ListedColormap(['g', 'r', 'b'])
grid_hat = classifier.predict(grid_test) # 预测分类值
grid_hat = grid_hat.reshape(x1.shape) # 使之与输入的形状相同
plt.pcolormesh(x1, x2, grid_hat, cmap = cm_light) # 预测值的显示
plt.scatter(data_x[:, 0], data_x[:, 1], c = label_y[:, 0], s = 30, cmap = cm_dark) # 样本
plt.scatter(test_data[:, 0], test_data[:, 1], c = test_label[:, 0], s = 30, edgecolors = 'k', zorder = 2,
cmap = cm_dark) # 圈中测试集样本点
plt.xlabel('花萼长度', fontsize = 13)
plt.ylabel('花萼宽度', fontsize = 13)
plt.xlim(x1_min, x1_max)
plt.ylim(x2_min, x2_max)
plt.title('基于SVM鸢尾花萼长宽度二特征分类')
plt.savefig('./Result/iris-sepal-cal.png')
plt.show()
def plotPetalShow(test_data, test_label, data_x, label_y, classifier):
x1_min, x1_max = data_x[:, 0].min(), data_x[:, 0].max() # 第2维特征的范围
x2_min, x2_max = data_x[:, 1].min(), data_x[:, 1].max() # 第3维特征的范围
x1, x2 = np.mgrid[x1_min:x1_max:200j, x2_min:x2_max:200j] # 生成网络采样点
# print(x1.shape) # (200, 200)
# print(x1.flat) # flat属性可以使得像遍历以一维数组的方法来遍历多维数组
grid_test = np.stack((x1.flat, x2.flat), axis = 1) # 测试点
# print(grid_test.shape) # (40000, 2)
# 指定默认字体
matplotlib.rcParams['font.sans-serif'] = ['SimHei']
# 设置颜色
cm_light = matplotlib.colors.ListedColormap(['#A0FFA0', '#FFA0A0', '#A0A0FF'])
cm_dark = matplotlib.colors.ListedColormap(['g', 'r', 'b'])
grid_hat = classifier.predict(grid_test) # 预测分类值
grid_hat = grid_hat.reshape(x1.shape) # 使之与输入的形状相同
plt.pcolormesh(x1, x2, grid_hat, cmap = cm_light) # 预测值的显示
plt.scatter(data_x[:, 0], data_x[:, 1], c = label_y[:, 0], s = 30, cmap = cm_dark) # 样本
plt.scatter(test_data[:, 0], test_data[:, 1], c = test_label[:, 0], s = 30, edgecolors = 'k', zorder = 2,
cmap = cm_dark) # 圈中测试集样本点
plt.xlabel('花瓣长度', fontsize = 13)
plt.ylabel('花瓣宽度', fontsize = 13)
plt.xlim(x1_min, x1_max)
plt.ylim(x2_min, x2_max)
plt.title('基于SVM鸢尾花瓣长宽度二特征分类')
plt.savefig('./Result/iris-petal-cla.png')
plt.show()
测试结果
进一步优化：
def Iris_Sepal_Petal_Cla(data_x, label_y):
data_x = np.stack((data_x[:, 0], data_x[:, 2], data_x[:, 3]), axis=1)
train_data, test_data, train_label, test_label = sklearn.model_selection.train_test_split(data_x, label_y,
random_state = 1,
train_size = 0.8,
test_size = 0.2)
classifier = svm.SVC(C = 0.5, kernel = 'linear', gamma = 10, decision_function_shape = 'ovr') # rbf
classifier.fit(train_data, train_label.ravel())
train_label_pre = classifier.predict(train_data)
test_label_pre = classifier.predict(test_data)
print('花萼和花瓣训练集acc:', accuracy_score(train_label, train_label_pre))
print('花萼和花瓣测试集acc:', accuracy_score(test_label, test_label_pre))
plotSPShow(classifier, data_x, label_y)
def plotSPShow(clf, x, y):
iris_feature = 'sepal length', 'sepal width', 'petal lenght', 'petal width'
# 开始画图
x0_min, x0_max = x[:, 0].min(), x[:, 0].max()
x1_min, x1_max = x[:, 1].min(), x[:, 1].max() # 第0列的范围
x2_min, x2_max = x[:, 2].min(), x[:, 2].max() # 第1列的范围
x0, x1, x2 = np.mgrid[x0_min:x0_max:50j, x1_min:x1_max:50j, x2_min:x2_max:50j] # 生成网格采样点,3D
grid_test = np.stack((x0.flat, x1.flat, x2.flat), axis=1) # stack():沿着新的轴加入一系列数组， flat的作用是将数组分解成可连续访问的元素,目的就是把他拉直后合并，并且不改变数组
print('grid_test:\n', grid_test)
grid_hat = clf.predict(grid_test) # 预测分类值 得到【0,0.。。。2,2,2】
print('grid_hat:\n', grid_hat)
grid_hat = grid_hat.reshape(x1.shape) # reshape grid_hat和x1形状一致
# 若3*3矩阵e，则e.shape()为3*3,表示3行3列
cm_light = matplotlib.colors.ListedColormap(['#A0FFA0', '#FFA0A0', '#A0A0FF'])
cm_dark = matplotlib.colors.ListedColormap(['g', 'r', 'b'])
fig = plt.figure()
ax = fig.add_subplot(projection='3d')
# s：marker标记的大小
# c: 颜色 可为单个，可为序列
# depthshade: 是否为散点标记着色以呈现深度外观。对 scatter() 的每次调用都将独立执行其深度着色。
# marker：样式
# alpha为点的透明度，在0~1之间
ax.scatter(xs=x1, ys=x2, zs=x0, zdir='z', s=10, c=grid_hat, depthshade=True, cmap=cm_light,alpha=0.01)
ax.scatter(xs=x[:,1], ys=x[:,2], zs=x[:,0], zdir='z', s=30, c=np.squeeze(y), depthshade=True, cmap=cm_dark, marker="^")
plt.title('基于SVM鸢尾花萼长度和花瓣长宽度三特征分类')
plt.savefig('./Result/iris-sepal-petal-cla.png')
plt.show()
if __name__ == "__main__":
txt_path = './/Result//SVM.txt'
data = np.loadtxt("./Data/iris.data", dtype = float, delimiter = ',', converters = {4: Iris_label})
# 基于SVM鸢尾花萼长宽度二特征分类
data_x, label_y = np.split(data, indices_or_sections = (4,), axis = 1) # x为数据，y为标签
Iris_Sepal_Petal_Cla(data_x, label_y)
Plot.py 将以上绘图.py代码汇总
# -*- coding: utf-8 -*- #
"""
@Project ：MachineLearningLesson
@File ：plot.py
@Author ：ZAY
@Time ：2023/6/5 21:41
@Annotation : " "
"""
# 确定坐标轴范围
import matplotlib
import matplotlib.pyplot as plt
import numpy as np
def plotSepalShow(test_data, test_label, data_x, label_y, classifier):
x1_min, x1_max = data_x[:, 0].min(), data_x[:, 0].max() # 第0维特征的范围
x2_min, x2_max = data_x[:, 1].min(), data_x[:, 1].max() # 第1维特征的范围
x1, x2 = np.mgrid[x1_min:x1_max:200j, x2_min:x2_max:200j] # 生成网络采样点
# print(x1.shape) # (200, 200)
# print(x1.flat) # flat属性可以使得像遍历以一维数组的方法来遍历多维数组
grid_test = np.stack((x1.flat, x2.flat), axis = 1) # 测试点
# print(grid_test.shape) # (40000, 2)
# 指定默认字体
matplotlib.rcParams['font.sans-serif'] = ['SimHei']
# 设置颜色
cm_light = matplotlib.colors.ListedColormap(['#A0FFA0', '#FFA0A0', '#A0A0FF'])
cm_dark = matplotlib.colors.ListedColormap(['g', 'r', 'b'])
grid_hat = classifier.predict(grid_test) # 预测分类值
grid_hat = grid_hat.reshape(x1.shape) # 使之与输入的形状相同
plt.pcolormesh(x1, x2, grid_hat, cmap = cm_light) # 预测值的显示
plt.scatter(data_x[:, 0], data_x[:, 1], c = label_y[:, 0], s = 30, cmap = cm_dark) # 样本
plt.scatter(test_data[:, 0], test_data[:, 1], c = test_label[:, 0], s = 30, edgecolors = 'k', zorder = 2,
cmap = cm_dark) # 圈中测试集样本点
plt.xlabel('花萼长度', fontsize = 13)
plt.ylabel('花萼宽度', fontsize = 13)
plt.xlim(x1_min, x1_max)
plt.ylim(x2_min, x2_max)
plt.title('基于SVM鸢尾花萼长宽度二特征分类')
plt.savefig('./Result/iris-sepal-cal.png')
plt.show()
def plotPetalShow(test_data, test_label, data_x, label_y, classifier):
x1_min, x1_max = data_x[:, 0].min(), data_x[:, 0].max() # 第2维特征的范围
x2_min, x2_max = data_x[:, 1].min(), data_x[:, 1].max() # 第3维特征的范围
x1, x2 = np.mgrid[x1_min:x1_max:200j, x2_min:x2_max:200j] # 生成网络采样点
# print(x1.shape) # (200, 200)
# print(x1.flat) # flat属性可以使得像遍历以一维数组的方法来遍历多维数组
grid_test = np.stack((x1.flat, x2.flat), axis = 1) # 测试点
# print(grid_test.shape) # (40000, 2)
# 指定默认字体
matplotlib.rcParams['font.sans-serif'] = ['SimHei']
# 设置颜色
cm_light = matplotlib.colors.ListedColormap(['#A0FFA0', '#FFA0A0', '#A0A0FF'])
cm_dark = matplotlib.colors.ListedColormap(['g', 'r', 'b'])
grid_hat = classifier.predict(grid_test) # 预测分类值
grid_hat = grid_hat.reshape(x1.shape) # 使之与输入的形状相同
plt.pcolormesh(x1, x2, grid_hat, cmap = cm_light) # 预测值的显示
plt.scatter(data_x[:, 0], data_x[:, 1], c = label_y[:, 0], s = 30, cmap = cm_dark) # 样本
plt.scatter(test_data[:, 0], test_data[:, 1], c = test_label[:, 0], s = 30, edgecolors = 'k', zorder = 2,
cmap = cm_dark) # 圈中测试集样本点
plt.xlabel('花瓣长度', fontsize = 13)
plt.ylabel('花瓣宽度', fontsize = 13)
plt.xlim(x1_min, x1_max)
plt.ylim(x2_min, x2_max)
plt.title('基于SVM鸢尾花瓣长宽度二特征分类')
plt.savefig('./Result/iris-petal-cla.png')
plt.show()
def plotSPShow(clf, x, y):
iris_feature = 'sepal length', 'sepal width', 'petal lenght', 'petal width'
# 开始画图
x0_min, x0_max = x[:, 0].min(), x[:, 0].max()
x1_min, x1_max = x[:, 1].min(), x[:, 1].max() # 第0列的范围
x2_min, x2_max = x[:, 2].min(), x[:, 2].max() # 第1列的范围
x0, x1, x2 = np.mgrid[x0_min:x0_max:50j, x1_min:x1_max:50j, x2_min:x2_max:50j] # 生成网格采样点,3D
grid_test = np.stack((x0.flat, x1.flat, x2.flat), axis=1) # stack():沿着新的轴加入一系列数组， flat的作用是将数组分解成可连续访问的元素,目的就是把他拉直后合并，并且不改变数组
print('grid_test:\n', grid_test)
grid_hat = clf.predict(grid_test) # 预测分类值 得到【0,0.。。。2,2,2】
print('grid_hat:\n', grid_hat)
grid_hat = grid_hat.reshape(x1.shape) # reshape grid_hat和x1形状一致
# 若3*3矩阵e，则e.shape()为3*3,表示3行3列
cm_light = matplotlib.colors.ListedColormap(['#A0FFA0', '#FFA0A0', '#A0A0FF'])
cm_dark = matplotlib.colors.ListedColormap(['g', 'r', 'b'])
fig = plt.figure()
ax = fig.add_subplot(projection='3d')
# s：marker标记的大小
# c: 颜色 可为单个，可为序列
# depthshade: 是否为散点标记着色以呈现深度外观。对 scatter() 的每次调用都将独立执行其深度着色。
# marker：样式
# alpha为点的透明度，在0~1之间
ax.scatter(xs=x1, ys=x2, zs=x0, zdir='z', s=10, c=grid_hat, depthshade=True, cmap=cm_light,alpha=0.01)
ax.scatter(xs=x[:,1], ys=x[:,2], zs=x[:,0], zdir='z', s=30, c=np.squeeze(y), depthshade=True, cmap=cm_dark, marker="^")
plt.title('基于SVM鸢尾花萼长度和花瓣长宽度三特征分类')
plt.savefig('./Result/iris-sepal-petal-cla.png')
plt.show()
测试结果
