机器学习02

1，回归算法-线性回归分析
线性回归详解
2，线性回归实例

# coding: gbk
from sklearn.datasets import fetch_california_housing
from sklearn.linear_model import LinearRegression,SGDRegressor
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import StandardScaler


def mylinear():
    """
    线性回归直接预测波士顿房价
    :return:
    """
    # 获取数据
    lb = fetch_california_housing()
    # 分割数据集到训练集和测试集
    x_train, x_test, y_train, y_test = train_test_split(lb.data, lb.target, test_size=0.25)
    # 进行标准化处理(特征值和目标值都需要标准化)
    std_x = StandardScaler()

    x_train = std_x.fit_transform(x_train)
    x_test = std_x.transform(x_test)

    # 目标值
    std_y = StandardScaler()

    y_train = std_y.fit_transform(y_train.reshape(-1, 1))
    y_test = std_y.transform(y_test.reshape(-1, 1))

    # estimator预测
    # 正规方程求解方式预测结果
    lr = LinearRegression()

    lr.fit(x_train,y_train)
    print(lr.coef_)

    # 预测测试集房价
    y_predict = std_y.inverse_transform(lr.predict(x_test))

    print("测试集里面每个房子的预测价格：", y_predict)

    # 梯度下降求解方式预测结果
    sgd = SGDRegressor()

    sgd.fit(x_train, y_train)
    print(sgd.coef_)

    # 预测测试集房价
    y_predict = std_y.inverse_transform(sgd.predict(x_test.reshape))

    print("sgd测试集里面每个房子的预测价格：", y_predict)


    return None


if __name__ == "__main__":
    mylinear()

3，回归性能评估
均方误差MSE:
API: mean_squared_error

4，分类算法-逻辑回归
1）模型的保存与加载

# 保存训练好的模型
    joblib.dump(lr,"B:/PycharmProjects/PythonProject/机器学习/model/test.pkl")

# 加载模型
    model = joblib.load("B:/PycharmProjects/PythonProject/机器学习/model/test.pkl")
    y_predict = std_y.inverse_transform(model.predict(x_test))

5，分类–逻辑回归实例
详解逻辑回归

# coding: gbk
import numpy as np
from sklearn.linear_model import LogisticRegression
import pandas as pd
from sklearn.metrics import classification_report
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import StandardScaler


def logistic():
    """
    逻辑回归
    :return:
    """
    # 构造列标签名字
    column = ['Sample code number', 'Clump Thickness', 'Unidormity of Cell Size', 'Uniformity of Cell Shape',
              'Marginal Adhesion', 'Single Epithelial Cell Size', 'Bare Nuclel', 'Bland Chromatin', 'Normal Nucleoli',
              'Mitoses', 'Class']

    # 读取数据
    data = pd.read_csv(
        "http://archive.ics.uci.edu/ml/machine-learning-databases/breast-cancer-wisconsin/breast-cancer-wisconsin.data",
        names=column)

    print(data)

    # 缺失值处理
    data = data.replace(to_replace='?', value=np.nan)
    data = data.dropna()

    # 进行数据的分割
    x_train, x_test, y_train, y_test = train_test_split(data[column[1:10]], data[column[10]], test_size=0.25)

    # j进行标准化处理
    std = StandardScaler()
    x_train = std.fit_transform(x_train)
    x_test - std.transform(x_test)

    # 逻辑回归的预测()
    lg = LogisticRegression(C=1.0)
    lg.fit(x_train, y_train)  # 进行训练，利用对数自然损值不断去求最小值，优化W值

    print(lg.coef_)  # 逻辑回归的权重参数
    y_predict = lg.predict(x_test)  # 获取预测值
    print("准确率：", lg.score(x_test, y_test))
    print("召回率：", classification_report(y_test, y_predict, labels=[2, 4], target_names=["良性", "恶性"]))

    return None


if __name__ == "__main__":
    logistic()

6，聚类算法k-means
详解k-means