机器学习之分类算法应用篇

清疚 2023-06-28 13:30 47阅读 0赞

和上一篇聚类算法类似，本文主要分享一个包装好的使用sklearn中常用分类器分类结果对比的代码，方便初步比较模型的时候使用，程序运行结果如下图所示。先给出代码，后续再做详细分析。

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzI2ODIyMDI5_size_16_color_FFFFFF_t_70][]

可执行Python3代码：

# 分类问题可运行代码
    from collections import Counter
    from sklearn.datasets import load_iris
    from imblearn.combine import SMOTEENN
    from imblearn.over_sampling import SMOTE
    from sklearn.preprocessing import StandardScaler, MinMaxScaler
    from sklearn.decomposition import PCA
    from sklearn.linear_model import LogisticRegression
    from sklearn.svm import SVC
    from sklearn.naive_bayes import GaussianNB
    from sklearn.tree import DecisionTreeClassifier
    from sklearn.ensemble import RandomForestClassifier, GradientBoostingClassifier
    from sklearn.model_selection import train_test_split, cross_val_predict
    from sklearn.metrics import classification_report
    from sklearn.pipeline import Pipeline
    import matplotlib.pyplot as plt
    import threading
    import warnings
    import numpy as np
    import time
    warnings.filterwarnings("ignore")
    
    model_list = [
        ("LR", LogisticRegression()),
        ("DT", DecisionTreeClassifier()),
        ("NB", GaussianNB()),
        ("RF", RandomForestClassifier()),
        ("GBDT", GradientBoostingClassifier())
    ]
    
    iris = load_iris()
    X, y = iris.data, iris.target
    
    # 上采样
    x_smo, y_smo = SMOTE(random_state=0).fit_sample(X, y)
    
    time_, acc_ = [], []
    for model in model_list:
        print("Now ", model[0], end=" ")
        pipe_lr = Pipeline([('st', MinMaxScaler()),
                            ('pca', PCA(n_components=0.99)),
                            ('clf', model[1])
                            ])
        t0 = time.time()
        y_pred = cross_val_predict(pipe_lr, x_smo, y_smo, cv=5)
        t1 = time.time()
        report = classification_report(y_smo, y_pred)
        print("({:.2f} s)".format(t1 - t0))
        print(report)
        time_.append(t1 - t0)
        acc_.append(float(report.split()[20]))
    
    fig, (ax1, ax2) = plt.subplots(1, 2)
    fig.set_size_inches(18, 7)
    
    ax1.bar(np.arange(len(time_)), time_)
    ax1.set_title("Time costed for each classifier (s)")
    ax1.set_xticks(range(0,len(time_), 1))
    ax1.set_xticklabels([model[0] for model in model_list])
    ax1.set_ylabel("Time(s)")
    
    ax2.bar(np.arange(len(time_)), acc_)
    ax2.set_title("Accuracy for each classifier (s)")
    ax2.set_xticks(range(0,len(time_), 1))
    ax2.set_xticklabels([model[0] for model in model_list])
    ax2.set_ylabel("Accuracy")
    
    plt.suptitle("Results on different classifiers", fontweight="bold", fontsize=15)
    plt.show()

（完）

[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzI2ODIyMDI5_size_16_color_FFFFFF_t_70]: https://img-blog.csdnimg.cn/20200109155532221.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzI2ODIyMDI5,size_16,color_FFFFFF,t_70