逻辑回归算法

import numpy as np
import matplotlib.pyplot as plt
from sklearn.datasets import load_iris
from sklearn.model_selection import train_test_split

class LogisticRegression:
    def __init__(self):
        self.coef_ = None
        self.intercept_ = None
        self._theta = None

    def _sigmoid(self, t):
        return 1 / (1 + np.exp(-t))

    def fit(self, x_train, y_trian, n):

        def J(theta, x_b, y):
            y_hat = self._sigmoid(x_b.dot(theta))

            return -np.sum(y * np.log(y_hat) + (1 - y) * np.log(1 - y_hat)) / len(y)

        def dJ(theta, x_b, y):
            return x_b.T.dot(self._sigmoid(x_b.dot(theta)) - y) / len(y)

        def gradient_descent(x_b, y, initial_theta, alpha=0.01, n_iters=1e4, epsilon=1e-8):
            theta = initial_theta
            iter = 0
            while iter < n_iters:
                gradient = dJ(theta, x_b, y)
                last_theta = theta
                theta = theta - alpha * gradient

                if abs(J(last_theta, x_b, y) - J(theta, x_b, y)) < epsilon:
                    break

                iter += 1
            return last_theta

        x_b = np.hstack([np.ones((len(x_train), 1)), x_train])
        initail_theta = np.zeros(x_b.shape[1])

        self._theta = gradient_descent(x_b, y_trian, initail_theta, n_iters=n)
        self.intercept_ = self._theta[0]
        self.coef_ = self._theta[1:]

    def predict_prob(self, x_test):
        x_b = np.hstack([np.ones((len(x_test), 1)), x_test])
        return self._sigmoid(x_b.dot(self._theta))

    def predict(self, x_test):
        prob = self.predict_prob(x_test)
        return np.array(prob >= 0.5,dtype='int')

    def score(self, x_test, y_test):
        return np.sum(y_test == self.predict(x_test)) / len(y_test)

# 鸢尾花数据集测试手写逻辑回归算法
iris = load_iris()
x = iris.data
y = iris.target

# 由于逻辑回归只能处理2分类问题，截取分类为0,1，保留x2个特征方便可视化
X = x[y < 2, :2]
Y = y[y < 2]

plt.scatter(X[Y == 0, 0], X[Y == 0, 1])
plt.scatter(X[Y == 1, 0], X[Y == 1, 1], color='r')
plt.show()

x_train, x_test, y_train, y_test = train_test_split(X, Y)

logistic = LogisticRegression()
logistic.fit(x_train, y_train, n=1e6)
y_predict = logistic.predict(x_test)
score = logistic.score(x_test, y_test)
print(score)
print(y_predict)
print(y_test)