Logistic回帰（勾配降下法）

logistic_regression_gradient_descent.py

import numpy as np

class LogisticRegressionGradientDescent:

    def __init__(self, eta=0.05, n_iter=100, random_state=1):
        self.eta = eta
        self.n_iter = n_iter
        self.random_state = random_state

    def fit(self, X, y):

        rgen = np.random.RandomState(self.random_state)
        _, n_features = X.shape
        self.w_ = rgen.normal(loc=0.0, scale=0.01, size = 1 + n_features)

        for i in range(self.n_iter):
            net_input = self.net_input(X)
            output = self.activation(net_input)
            errors = y - output

            self.w_[1:] += self.eta * np.dot(X.T, errors)
            self.w_[0]  += self.eta * errors.sum()

        return self

    def net_input(self, X):
        return np.dot(X, self.w_[1:]) + self.w_[0]

    def activation(self, z):
        """
        ロジスティックシグモイド関数による活性化関数
        """
        return 1.0 / (1.0 + np.exp(-np.clip(z, -250, 250)))

    def predict(self, X):
        return np.where(self.activation(self.net_input(X)) >= 0.5, 1, 0)
        # 以下でも同じ結果
        # return np.where(self.net_input(X) >= 0.0, 1, 0)