Python机器学习经典模型实战篇

模型实战篇"/>

Python机器学习经典模型实战篇

B站账号@狼群里的小杨，记得点赞收藏加关注，一键三连哦！

朴素贝叶斯算法实例文本分类

from sklearn.datasets import  fetch_20newsgroups
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.model_selection import train_test_split
from sklearn.naive_bayes import  MultinomialNB
from sklearn.metrics import classification_reportdef naviebayes():"""朴素贝叶斯进行文本分类:return: None"""# 获取新闻数据集news = fetch_20newsgroups(subset='all')# 进行数据分割x_train,x_test,y_train,y_test = train_test_split(news.data, news.target, test_size=0.2)# 对数据集进行特征抽取tf = TfidfVectorizer()# 以训练集当中的词的列表进行每篇文章重要性统计['a','b','c','d',]训练值必须是已经分好词的数据，将分词结果用空格连接起来形成每一篇文章的数据。x_train = tf.fit_transform(x_train)print(tf.get_feature_names())x_test = tf.transform(x_test)# 进行朴素贝叶斯算法的预测mlt = MultinomialNB(alpha=1.0)print(x_train)#设置训练集mlt.fit(x_train, y_train)y_predict = mlt.predict(x_test)print("预测的文章类别为: ",y_predict)# 得出准确率print("准确率：",mlt.score(x_test,y_test))print("每个类别的精确率和召回率：\n",classification_report(y_test, y_predict, target_names=news.target_names))return Noneif __name__ =="__main__":naviebayes()

决策树实例-预测泰坦尼克号乘客生存率

import pandas as pd
from sklearn.feature_extraction import DictVectorizer #导入特征抽取的函数库
from sklearn.model_selection import train_test_split # 导入数据分割模块
from sklearn.tree import DecisionTreeClassifier # 导入决策树模型库
from sklearn.tree import export_graphviz #导入生成决策树图片的库def decision():"""决策树对泰坦尼克号进行预测生死:return: None"""# 获取数据# titan = pd.read_csv('titan.csv')titan = pd.read_csv(".txt")# 处理数据，找出特征值和目标值x = titan[['pclass', 'age', 'sex']]y= titan['survived']print(x)#  缺失值处理x['age'].fillna(x['age'].mean(), inplace=True)# 分割数据集到训练集合测试集x_train, x_test, y_train, y_test = train_test_split(x, y, test_size=0.25)#  进行处理(特征工程) 特征-> 类别-> one_hot编码dict = DictVectorizer(sparse=False)x_train = dict.fit_transform(x_train.to_dict(orient="records")) #将每一行的数据转换成字典列表。 转换成字典才可以进行特征抽取print(dict.get_feature_names())x_test = dict.transform(x_test.to_dict(orient="records"))# print(x_train)# 用决策树进行预测dec = DecisionTreeClassifier(max_depth=8) # 传入最大深度值dec.fit(x_train, y_train)# 预测准确率print("预测的准确率: ", dec.score(x_test,y_test))# 导出决策树的结构export_graphviz(dec, out_file="./tree.dot",feature_names=['年龄','pclass=1st', 'pclass=2nd', 'pclass=3rd', '女性', '男性'])return Noneif __name__ =="__main__":decision()

随机森林实例-泰坦尼克号乘客生存预测

import pandas as pd
from sklearn.feature_extraction import DictVectorizer #导入特征抽取的函数库
from sklearn.model_selection import train_test_split # 导入数据分割模块
from sklearn.tree import DecisionTreeClassifier # 导入决策树模型库
from sklearn.tree import export_graphviz #导入生成决策树图片的库
from sklearn.ensemble import RandomForestClassifier # 导入随机森林算法库
from sklearn.model_selection import GridSearchCV # 导入网格搜索与交叉验证库def decision():"""决策树对泰坦尼克号进行预测生死:return: None"""# 获取数据# titan = pd.read_csv('titan.csv')titan = pd.read_csv(".txt")# 处理数据，找出特征值和目标值x = titan[['pclass', 'age', 'sex']]y= titan['survived']print(x)#  缺失值处理x['age'].fillna(x['age'].mean(), inplace=True)# 分割数据集到训练集合测试集x_train, x_test, y_train, y_test = train_test_split(x, y, test_size=0.25)#  进行处理(特征工程) 特征-> 类别-> one_hot编码dict = DictVectorizer(sparse=False)x_train = dict.fit_transform(x_train.to_dict(orient="records")) # 转换成字典才可以进行特征抽取print(dict.get_feature_names())x_test = dict.transform(x_test.to_dict(orient="records"))# print(x_train)#用决策树进行预测# dec = DecisionTreeClassifier(max_depth=8) # 传入最大深度值## dec.fit(x_train, y_train)## # 预测准确率# print("预测的准确率: ", dec.score(x_test,y_test))## # 导出决策树的结构# export_graphviz(dec, out_file="./tree.dot",feature_names=['年龄','pclass=1st', 'pclass=2nd', 'pclass=3rd', '女性', '男性'])# 随机森林进行预测（超参数调优）rf = RandomForestClassifier()# 设置参数字典，第一个参数列表包含决策树的数目，第二个参数列表是随机数深度，从列表中选取一个值两辆组合进行计算。# 这里共有6x5=30种组合情况需要计算param = {"n_estimators": [120, 200, 300, 500, 800, 1200], "max_depth":[5, 8, 15, 25,30]}# 网格搜索与交叉验证gc = GridSearchCV(rf, param_grid=param, cv=2)#可以计算信息熵的大小gc.fit(x_train, y_train)print("准确率：",gc.score(x_test, y_test))print("查看选择的参数模型：",gc.best_params_)#这里不能导出随机森林的结构，只能导出单个决策树的结构生成图片return Noneif __name__ =="__main__":decision()

线性回归、梯度下降、岭回归预测波士顿房价

from sklearn.datasets import load_boston # 获取波士顿的房价数据
from  sklearn.linear_model import LinearRegression,SGDRegressor,Ridge # 导入线性回归,梯度下降，岭回归的函数库
from sklearn.model_selection import train_test_split # 导入数据集分割函数
from sklearn.preprocessing import StandardScaler # 导入标准化函数库
from sklearn.metrics import mean_squared_error #回归评估APIdef mylinear():"""线性回归直接预测房子价格:return:None"""# 获取数据lb = load_boston()#分割数据集到训练集合测试集x_train,x_test, y_train,y_test =   train_test_split(lb.data, lb.target,test_size=0.25)print(y_train, y_test)#进行标准化处理(?) 目标值处理？#特征值和目标值是都必须进行标准化处理，实例化两个标准化APIstd_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_)# 打印权重参数#预测测试集的房子价格y_lr_predict = std_y.inverse_transform(lr.predict(x_test))print("正规方程测试集里面每个房子的预测价格：\n",y_lr_predict)print("正规方程的均方误差：",mean_squared_error(std_y.inverse_transform(y_test), y_lr_predict))# 梯度下降去进行房价预测sgd = SGDRegressor()sgd.fit(x_train, y_train)print(sgd.coef_)  # 打印权重参数# 预测测试集的房子价格y_sgd_predict = std_y.inverse_transform(sgd.predict(x_test))print("梯度下降测试集里面每个房子的预测价格：\n", y_sgd_predict)print("梯度下降的均方误差：", mean_squared_error(std_y.inverse_transform(y_test), y_sgd_predict))# 岭回归去进行房价预测rd = Ridge(alpha=1.0)rd.fit(x_train, y_train)print(rd.coef_)  # 打印权重参数# 预测测试集的房子价格y_rd_predict = std_y.inverse_transform(rd.predict(x_test))print("岭回归测试集里面每个房子的预测价格：\n", y_rd_predict)print("岭回归的均方误差：", mean_squared_error(std_y.inverse_transform(y_test), y_rd_predict))return Noneif __name__=="__main__":mylinear()

保存和加载API

from sklearn.datasets import load_boston # 获取波士顿的房价数据
from  sklearn.linear_model import LinearRegression,SGDRegressor,Ridge # 导入线性回归,梯度下降，岭回归的函数库
from sklearn.model_selection import train_test_split # 导入数据集分割函数
from sklearn.preprocessing import StandardScaler # 导入标准化函数库
from sklearn.externals import joblib #导入保存模型的APIdef mylinear():"""线性回归直接预测房子价格:return:None"""# 获取数据lb = load_boston()#分割数据集到训练集合测试集x_train,x_test, y_train,y_test =   train_test_split(lb.data, lb.target,test_size=0.25)print(y_train, y_test)#进行标准化处理(?) 目标值处理？#特征值和目标值是都必须进行标准化处理，实例化两个标准化APIstd_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))# 预测房价结果# model = joblib.load("./test.pkl") # 加载之前保存的模型，当前文件夹下面的test.pkl文件即为之前保存的模型# # y_predict = std_y.inverse_transform(model.predict(x_test))# print("保存的模型预测的结果：\n",y_predict)#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_)# 打印权重参数#保存训练好的模型到当前文件夹下面的test.pkl文件中joblib.dump(lr, "./test.pkl")return Noneif __name__=="__main__":mylinear()

逻辑回归二分类实例癌症预测

import numpy as np
import pandas as pd
from sklearn.model_selection import train_test_split # 导入数据集分割函数库
from sklearn.preprocessing import StandardScaler # 导入标准化的API
from  sklearn.linear_model import LogisticRegression # 导入逻辑回归API
from sklearn.metrics import classification_report #分类模型的评估APIdef logistic():"""逻辑回归做二分类进行癌症预测(根据细胞的属性特征):return: None"""# 构造列标签名字column = ['Sample code number', 'Clump Thickness', 'Uniformity of Cell Size','Uniformity of Cell Shape','Marginal Adhesion','Single Epithelial Cell Size', 'Bare Nuclei','Bland Chromatin','Normal Nucleoli', 'Mitoses', 'Class']#读取数据data = pd.read_csv(".data", names=column)print(data)# 缺失值进行处理data = data.replace(to_replace='?', value=np.nan)data = data.dropna()#进行数据分割x_train, x_test,y_train, y_test = train_test_split(data[column[1:10]], data[column[10]], test_size=0.25)print(y_train, y_test)#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lg = LogisticRegression(C=1.0)lg.fit(x_train, y_train)print(lg.coef_)y_predict = lg.predict(x_test)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 Noneif __name__=="__main__":logistic()

聚类Kmeans算法实现市场菜篮子案例

import pandas as pd
import matplotlib.pyplot as plt
from sklearn.decomposition import PCA
from sklearn.cluster import KMeans # 导入聚类的API# 读取四张表的数据
prior = pd.read_csv("order_products__prior.csv")
products = pd.read_csv("products.csv")
orders = pd.read_csv("orders.csv")
aisles = pd.read_csv("aisles.csv")# 合并四张表到同一张表（用户-物品类别）
_mg = pd.merge(prior, products, on=['product_id', 'product_id'])
_mg = pd.merge(_mg, orders, on=['order_id','order_id'])
mt = pd.merge(_mg, aisles, on=['aisle_id','aisle_id'])
print(mt.head(10))#交叉表（特殊的分组工具）
#按user_id分类，统计各个分类中aisle的频数
cross = pd.crosstab(mt['user_id'], mt['aisle'])
print(cross.head(10))# 进行主成分分析
pca = PCA(n_components=0.9)
data = pca.fit_transform(cross)# 把样本数量减少
x = data[:500]
print(x.shape)km =KMeans(n_clusters=4)
km.fit(x)predict = km.predict(x)
print("预测结果：\n",predict)# 显示聚类的结果
plt.figure(figsize=(6, 6))# 指定画板大小#建立四个颜色的列表
colored = ['orange', 'green', 'blue', 'purple']
colr = [colored[i] for i in predict]
#绘制散点图
plt.scatter(x[:,1], x[:, 20], color=colr)
plt.xlabel("1")
plt.ylabel("20")
plt.show()# print(data.shape)