以类方法定义Adaboost
import numpy as np
import pandas as pd
from sklearn.tree import DecisionTreeClassifier
from sklearn.metrics import f1_score, accuracy_score
class Adaboost:
"""
__init__(self, num_classifiers): 构造方法，用于初始化Adaboost对象。num_classifiers指定要使用的分类器数量。
weighted_error(y_true, y_pred, weights): 静态方法，计算加权错误率。它接受真实标签y_true、预测标签y_pred和样本权重weights作为输入，并返回加权错误率。
update_weights(weights, alpha, y_true, y_pred): 静态方法，根据分类器权重更新样本权重。它接受当前样本权重weights、分类器权重alpha、真实标签y_true和预测标签y_pred作为输入，并返回更新后的样本权重。
fit(self, X, y): 训练Adaboost分类器。它接受训练数据X和标签y作为输入，并使用num_classifiers个决策树分类器进行训练
"""
def __init__(self, num_classifiers):
self.num_classifiers = num_classifiers
self.classifiers = []
self.classifier_weights = []
@staticmethod
def weighted_error(y_true, y_pred, weights):
return np.sum(weights[y_true != y_pred])
@staticmethod
def update_weights(weights, alpha, y_true, y_pred):
weights_new = weights * np.exp(alpha * (y_true != y_pred))
return weights_new / np.sum(weights_new)
def fit(self, X, y):
# 初始化样本权重
sample_weights = np.ones(len(X)) / len(X)
for _ in range(self.num_classifiers):
# 创建分类器并训练
classifier = DecisionTreeClassifier(max_depth=1)
classifier.fit(X, y, sample_weight=sample_weights)
# 预测
y_pred = classifier.predict(X)
# 计算错误率和分类器权重
error = self.weighted_error(y, y_pred, sample_weights)
print(error)
print('-'* 50)
alpha = 0.5 * np.log((1 - error) / error)
# 更新样本权重
sample_weights = self.update_weights(sample_weights, alpha, y, y_pred)
# 保存分类器和权重
self.classifiers.append(classifier)
self.classifier_weights.append(alpha)
def predict(self, X):
# 初始化预测结果
y_pred = np.zeros(len(X))
for classifier, alpha in zip(self.classifiers, self.classifier_weights):
# 对每个分类器进行预测
y_pred_single = classifier.predict(X)
# 根据权重进行加权
y_pred += alpha * y_pred_single
# 使用符号函数将预测结果转换为类标签
y_pred = np.sign(y_pred)
print("预测y_pred:", y_pred)
return y_pred
# 准备数据
df = pd.DataFrame([[0, 1],
[1, 1],
[2, 1],
[3, -1],
[4, -1],
[5, -1],
[6, 1],
[7, 1],
[8, 1],
[9, -1]])
# 提取数据信息
X = df.iloc[:, :-1]
Y = df.iloc[:, -1]
print(X)
print(Y)
# 实例化类对象
A = Adaboost(20)
A.fit(X, Y)
A.predict(X)
E:/XXXXX/机器学习课程/集成学习/Cls_adaboost.py
0
0 0
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
0 1
1 1
2 1
3 -1
4 -1
5 -1
6 1
7 1
8 1
9 -1
Name: 1, dtype: int64
0.30000000000000004
--------------------------------------------------
0.34534632929202286
--------------------------------------------------
0.35008562397019116
--------------------------------------------------
0.28001073918680375
--------------------------------------------------
0.25381778971648017
--------------------------------------------------
0.3064962299371764
--------------------------------------------------
0.2816048333195814
--------------------------------------------------
0.2731345864165943
--------------------------------------------------
0.29156133793693806
--------------------------------------------------
0.2823934402541752
--------------------------------------------------
0.33466386078918664
--------------------------------------------------
0.24608157437902595
--------------------------------------------------
0.3502647590291356
--------------------------------------------------
0.25394656115341074
--------------------------------------------------
0.2731513767570225
--------------------------------------------------
0.3056825853100197
--------------------------------------------------
0.2720919041411729
--------------------------------------------------
0.34005542241591524
--------------------------------------------------
0.24744912020573245
--------------------------------------------------
0.3529066226801387
--------------------------------------------------
预测y_pred: [ 1. 1. 1. -1. -1. -1. 1. 1. 1. -1.]
Process finished with exit code 0