最近几天学习了机器学习经典算法，通过此次学习入门了机器学习，并将经典算法的代码实现并记录下来，方便后续查找与使用。
这次记录主要分为两部分：第一部分是机器学习思维导图，以框架的形式描述机器学习开发流程，并附有相关的具体python库，做索引使用；第二部分是相关算法的代码实现（其实就是调包），方便后面使用时直接复制粘贴，改改就可以用，尤其是在数学建模中很实用。

第一部分，思维导图：

第二部分，代码示例：

机器学习代码示例

导包

import numpy as np
import pandas as pd
from matplotlib.pyplot import plot as plt

from sklearn.datasets import load_iris
from sklearn.model_selection import train_test_split
from sklearn.feature_extraction import DictVectorizer
from sklearn.feature_extraction.text import CountVectorizer
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.preprocessing import MinMaxScaler
from sklearn.preprocessing import StandardScaler
from sklearn.feature_selection import VarianceThreshold
from scipy.stats import pearsonr

from sklearn.model_selection import GridSearchCV
from sklearn.neighbors import KNeighborsClassifier
from sklearn.naive_bayes import MultinomialNB
from sklearn.tree import DecisionTreeClassifier, export_graphviz
from sklearn.ensemble import RandomForestClassifier

from sklearn.linear_model import LinearRegression, SGDRegressor, Ridge, LogisticRegression
from sklearn.metrics import mean_squared_error
from sklearn.metrics import classification_report
from sklearn.metrics import roc_auc_score
from sklearn.cluster import KMeans
from sklearn.metrics import silhouette_score
import joblib

特征工程

特征抽取

def dict_demo():
    data = [{'city': '北京', 'temperature': 100}, {'city': '上海', 'temperature': 200},
            {'city': '广州', 'temperature': 300}]
    transfer = DictVectorizer()
    data_new = transfer.fit_transform(data)
    data_new = data_new.toarray()
    print(data_new)
    print(transfer.get_feature_names_out())


# dict_demo()

def count_demo():
    data = ["I love love China", "I don't love China"]
    transfer = CountVectorizer()
    data_new = transfer.fit_transform(data)
    data_new = data_new.toarray()
    print(data_new)
    print(transfer.get_feature_names_out())


# count_demo()

def chinese_demo(d):
    tt = " ".join(list(jieba.cut(d)))
    return tt


# data = [
#     "晚风轻轻飘荡,心事都不去想,那失望也不失望,惆怅也不惆怅,都在风中飞扬",
#     "晚风轻轻飘荡,随我迎波逐浪,那欢畅都更欢畅,幻想更幻想,就像 你还在身旁"]
# res = []
# for t in data:
#     res.append(chinese_demo(t))
#
# transfer = TfidfVectorizer()
# new_data = transfer.fit_transform(res)
# new_data = new_data.toarray()
# print(new_data)
# print(transfer.get_feature_names_out())

数据预处理

def minmax_demo():
    data = pd.read_csv("datasets/dating.txt")
    data = data.iloc[:, 0:3]
    print(data)
    transfer = MinMaxScaler()
    data_new = transfer.fit_transform(data)
    print(data_new)
    return None


# minmax_demo()


def standard_demo():
    data = pd.read_csv("datasets/dating.txt")
    data = data.iloc[:, 0:3]
    print(data)
    transfer = StandardScaler()
    data_new = transfer.fit_transform(data)
    print(data_new)
    return None


# standard_demo()

def stats_demo():
    data = pd.read_csv("./datasets/factor_returns.csv")
    data = data.iloc[:, 1:10]
    transfer = VarianceThreshold(threshold=10)
    data_new = transfer.fit_transform(data)
    print(data_new)
    print(data_new.shape)
    df = pd.DataFrame(data_new, columns=transfer.get_feature_names_out())
    print(df)


# stats_demo()

def pear_demo():
    data = pd.read_csv("./datasets/factor_returns.csv")
    data = data.iloc[:, 1:10]
    print(data.corr(method="pearson"))


# pear_demo()

模型训练

分类算法

KNN

# 读取数据
iris = load_iris()
# 数据集划分
x_train, x_test, y_train, y_test = train_test_split(iris.data, iris.target, test_size=0.3, random_state=42)
# 数据标准化
transfer = StandardScaler()
transfer.fit(x_train)
x_train = transfer.transform(x_train)
x_test = transfer.transform(x_test)
# 模型训练
estimator = KNeighborsClassifier(n_neighbors=i)
estimator.fit(x_train, y_train)
# 模型预测
y_predict = estimator.predict(x_test)
score = estimator.score(x_test, y_test)
print("score:", score)

朴素贝叶斯

new = fetch_20newsgroups(subset="all")
x_train, x_test, y_train, y_test = train_test_split(new.data, new.target, random_state=42)
# 文本特征提取
transfer = TfidfVectorizer()
transfer.fit(x_train)
x_train = transfer.transform(x_train)
x_test = transfer.transform(x_test)
estimator = MultinomialNB()
estimator.fit(x_train, y_train)
score = estimator.score(x_test, y_test)
print(score)

决策树

iris = load_iris()
x_train, x_test, y_train, y_test = train_test_split(iris.data, iris.target, test_size=0.3, random_state=42)
estimator = DecisionTreeClassifier(criterion='gini')
estimator.fit(x_train, y_train)
score = estimator.score(x_test, y_test)
print(score)
# 决策树可视化
export_graphviz(estimator, out_file='tree.dot', feature_names=iris.feature_names)

随机森林

x_train, x_test, y_train, y_test = train_test_split(x, y, train_size=0.7)
estimator = RandomForestClassifier(random_state=42, max_features='sqrt')
param_dict = {'n_estimators': range(10, 50), 'max_depth': range(5, 10)}
estimator = GridSearchCV(estimator=estimator, param_grid=param_dict, cv=3)
estimator.fit(x_train, y_train)
print(estimator.best_score_)
print(estimator.best_estimator_)
print(estimator.best_params_)

回归算法

线性回归

def demo1():
    data_url = "http://lib.stat.cmu.edu/datasets/boston"
    raw_df = pd.read_csv(data_url, sep="\s+", skiprows=22, header=None)
    data = np.hstack([raw_df.values[::2, :], raw_df.values[1::2, :2]])
    target = raw_df.values[1::2, 2]
    x_train, x_test, y_train, y_test = train_test_split(data, target, train_size=0.7, random_state=42)
    transfer = StandardScaler()
    transfer.fit(x_train)
    x_train = transfer.transform(x_train)
    x_test = transfer.transform(x_test)
    estimator = LinearRegression()
    estimator.fit(x_train, y_train)
    y_predict = estimator.predict(x_test)
    mse = mean_squared_error(y_test, y_predict)
    print("正规方程-", estimator.coef_)
    print("正规方程-", estimator.intercept_)
    print(mse)


def demo2():
    data_url = "http://lib.stat.cmu.edu/datasets/boston"
    raw_df = pd.read_csv(data_url, sep="\s+", skiprows=22, header=None)
    data = np.hstack([raw_df.values[::2, :], raw_df.values[1::2, :2]])
    target = raw_df.values[1::2, 2]
    x_train, x_test, y_train, y_test = train_test_split(data, target, train_size=0.7, random_state=42)
    transfer = StandardScaler()
    transfer.fit(x_train)
    x_train = transfer.transform(x_train)
    x_test = transfer.transform(x_test)
    estimator = SGDRegressor()
    estimator.fit(x_train, y_train)
    y_predict = estimator.predict(x_test)
    mse = mean_squared_error(y_test, y_predict)
    print("梯度下降", estimator.coef_)
    print("梯度下降", estimator.intercept_)
    print(mse)

岭回归

def demo3():
    data_url = "http://lib.stat.cmu.edu/datasets/boston"
    raw_df = pd.read_csv(data_url, sep="\s+", skiprows=22, header=None)
    data = np.hstack([raw_df.values[::2, :], raw_df.values[1::2, :2]])
    target = raw_df.values[1::2, 2]
    x_train, x_test, y_train, y_test = train_test_split(data, target, train_size=0.7, random_state=42)
    transfer = StandardScaler()
    transfer.fit(x_train)
    x_train = transfer.transform(x_train)
    x_test = transfer.transform(x_test)
    estimator = Ridge()
    estimator.fit(x_train, y_train)
    y_predict = estimator.predict(x_test)
    mse = mean_squared_error(y_test, y_predict)
    print("梯度下降", estimator.coef_)
    print("梯度下降", estimator.intercept_)
    print(mse)

逻辑回归

def demo4():
    data = pd.read_csv("./datasets/breast-cancer-wisconsin.data",
                       names=['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'])
    data.replace(to_replace="?", value=np.nan, inplace=True)
    data.dropna(inplace=True)
    x = data.iloc[:, 1:-1]
    y = data['Class']
    x_train, x_test, y_train, y_test = train_test_split(x, y, train_size=0.7, random_state=42)
    transfer = StandardScaler()
    transfer.fit(x_train)
    x_train = transfer.transform(x_train)
    x_test = transfer.transform(x_test)
    estimator = LogisticRegression()
    estimator.fit(x_train, y_train)
    # joblib.dump(estimator, 'estimator.pkl')
    # estimator = joblib.load('estimator.pkl')
    y_predict = estimator.predict(x_test)
    print(estimator.coef_)
    print(estimator.intercept_)
    score = estimator.score(x_test, y_test)
    print(score)
    report = classification_report(y_test, y_predict, labels=[2, 4], target_names=["良性", "恶性"])
    print(report)
    auc = roc_auc_score(y_test, y_predict)
    print(auc)

聚类算法

KMeans

data = pd.read_csv("./datasets/factor_returns.csv")
data = data.iloc[:, 1:10]
transfer = VarianceThreshold(threshold=10)
data_new = transfer.fit_transform(data)
# df = pd.DataFrame(data_new, columns=transfer.get_feature_names_out())
estimator = KMeans()
estimator.fit(data_new)
y_predict = estimator.predict(data_new)
print(y_predict)
s = silhouette_score(data_new, y_predict)
print(s)

模型调优

# 网格搜索与交叉验证：以KNN为例
iris = load_iris()
x_train, x_test, y_train, y_test = train_test_split(iris.data, iris.target, test_size=0.3, random_state=42)
transfer = StandardScaler()
transfer.fit(x_train)
x_train = transfer.transform(x_train)
x_test = transfer.transform(x_test)
estimator = KNeighborsClassifier()
# 网格搜素设置
para_dict = {"n_neighbors": range(1, 10)}
estimator = GridSearchCV(estimator, para_dict, cv=10)
estimator.fit(x_train, y_train)
# 最佳参数
print("best_score_:", estimator.best_score_)
print("best_estimator_:", estimator.best_estimator_)
print("best_params_:", estimator.best_params_)

本文作者：CodingOrange
本文链接：https://www.cnblogs.com/CodingOrange/p/17642747.html
转载请注明出处！