[코드 구현] logistic regression

numpy 로 구현

class LogisticRegression:
    def __init__(self, lr=0.01, num_iter=1000, fit_intercept=True, verbose=False):
        self.lr = lr
        self.num_iter = num_iter
        self.fit_intercept = fit_intercept
        self.verbose = verbose
        self.eps = 1e-10
        self.threshold = 0.5
        self.loss_history = list()
    
    def __add_intercept(self, X):
        intercept = np.ones((X.shape[0], 1))
        return np.concatenate((intercept, X), axis=1)
    
    def __sigmoid(self, z):
        return 1 / (1 + np.exp(-z))
        
    def __loss(self, h, y):
        return (-y * np.log(h+ self.eps) - (1 - y) * np.log(1 - h + self.eps)).mean()
    
    def fit(self, X, y):
        if self.fit_intercept:
            X = self.__add_intercept(X)
        
        # weights initialization
        self.theta = np.zeros(X.shape[1])
        
        for i in range(self.num_iter):
            logit = np.dot(X, self.theta)
            hypothesis = self.__sigmoid(logit)
            gradient = np.dot(X.T, (hypothesis - y)) / y.size
            self.theta -= self.lr * gradient
            
            if self.verbose == True and i % 10 == 0:
                loss = self.__loss(hypothesis, y)
                print(f'epoch: {i} \t loss: {loss} \t')
                self.loss_history.append(loss)
        return self.loss_history
    
    def predict_prob(self, X):
        if self.fit_intercept:
            X = self.__add_intercept(X)
    
        return self.__sigmoid(np.dot(X, self.theta))
    
    def predict(self, X):
        predicted_labels = np.where(self.predict_prob(X) > self.threshold, 1, 0)
        return predicted_labels
 
    def eval(self, x, y):
        res_y = np.round(self.predict_prob(x), 0)
        accuracy = np.sum(res_y==y) / len(y)
 
        return accuracy