[人工智能]机器学习_B站黑马程序员_学习1

1KNN算法

2朴素贝叶斯算法

题目：20类新闻分类

from sklearn.datasets import fetch_20newsgroups
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.naive_bayes import MultinomialNB
from sklearn.model_selection import train_test_split


def nb_news():
    # 1）获取数据
    news = fetch_20newsgroups(subset='all')

    # 2）划分数据集
    x_train, x_test, y_train, y_test = train_test_split(news.data, news.target)

    # 3）文本特征抽取
    transfer = TfidfVectorizer()
    x_train = transfer.fit_transform(x_train)
    x_test = transfer.transform(x_test)

    # 4）朴素贝叶斯算法预估
    estimator = MultinomialNB()
    estimator.fit(x_train, y_train)

    # 5)模型评估
    # 方法1：直接比对真实值和预测值
    y_predict = estimator.predict(x_test)
    print('y_predict\n', y_predict)
    print('真实值和预测值对比\n', y_predict == y_test)
    # 方法2：计算准确率
    score = estimator.score(x_test, y_test)
    print('准确率为\n', score)
    return None


if __name__ == '__main__':
    nb_news()

结果