头歌机器学习 — Adaboost第二关：Adaboost算法
```python
# encoding=utf8
import numpy as np
from sklearn.tree import DecisionTreeClassifier
from sklearn.ensemble import AdaBoostClassifier# adaboost算法
class AdaBoost:'''input:n_estimators(int):迭代轮数learning_rate(float):弱分类器权重缩减系数'''def __init__(self, n_estimators=50, learning_rate=1.0):self.clf_num = n_estimatorsself.learning_rate = learning_ratedef init_args(self, datasets, labels):self.X = datasetsself.Y = labelsself.M, self.N = datasets.shape# 弱分类器数目和集合self.clf_sets = []# 初始化weightsself.weights = [1.0 / self.M] * self.M# G(x)系数 alphaself.alpha = []# ********* Begin *********#def _G(self, features, labels, weights):'''input:features(ndarray):数据特征labels(ndarray):数据标签weights(ndarray):样本权重系数'''e = 0for i in range(weights.shape[0]):if (labels[i] == self.G(self.X[i], self.clif_sets, self.alpha)):e += weights[i]return e# 计算alphadef _alpha(self, error):return 0.5 * np.log((1 - error) / error)# 规范化因子def _Z(self, weights, a, clf):return np.sum(weights * np.exp(-a * self.Y * self.G(self.X, clf, self.alpha)))# 权值更新def _w(self, a, clf, Z):w = np.zeros(self.weights.shape)for i in range(self.M):w[i] = weights[i] * np.exp(-a * self.Y[i] * G(x, clf, self.alpha)) / Zself.weights = w# G(x)的线性组合def G(self, x, v, direct):result = 0x = x.reshape(1, -1)for i in range(len(v)):result += v[i].predict(x) * direct[i]return resultdef fit(self, X, y):'''X(ndarray):训练数据y(ndarray):训练标签'''# 计算G(x)系数aself.init_args(X, y)'''for i in range(100):classifier = DecisionTreeClassifier(max_depth=3)classifier.fit(X, y)self.clf_sets.append(classifier)e = 0for i in range(len(self.weights)):temp = -1if classifier.predict(X[i].reshape(1,-1))>0:temp = 1if(self.Y[i] == temp):e += self.weights[i]a = self._alpha(e)self.alpha.append(a)z = self._Z(self.weights, a, self.clf_sets)self._w(a, self.clf_sets, z)'''# 记录分类器# 规范化因子# 权值更新def predict(self, data):'''input:data(ndarray):单个样本output:预测为正样本返回+1，负样本返回-1'''ada = AdaBoostClassifier(n_estimators=100, learning_rate=0.1)ada.fit(self.X, self.Y)data = data.reshape(1, -1)predict = ada.predict(data)return predict[0]# ********* End *********#