设计并实现基于Python的朴素贝叶斯算法

以下是基于Python的朴素贝叶斯算法的实现：

import math

class NaiveBayes:
    def __init__(self):
        self.classes = None
        self.class_probabilities = None
        self.word_counts = None
        self.word_probabilities = None

    def fit(self, X, y):
        self.classes = set(y)
        self.class_probabilities = {}
        self.word_counts = {}
        self.word_probabilities = {}

        # 计算每个类别的概率
        for c in self.classes:
            self.class_probabilities[c] = y.count(c) / len(y)

        # 计算每个单词在每个类别中出现的次数
        for i in range(len(X)):
            for word in X[i].split():
                if word not in self.word_counts:
                    self.word_counts[word] = {}
                    for c in self.classes:
                        self.word_counts[word][c] = 0
                self.word_counts[word][y[i]] += 1

        # 计算每个单词在每个类别中出现的概率
        for word in self.word_counts:
            self.word_probabilities[word] = {}
            for c in self.classes:
                self.word_probabilities[word][c] = (self.word_counts[word][c] + 1) / (sum(self.word_counts[word].values()) + len(self.word_counts))

    def predict(self, X):
        result = []
        for x in X:
            probabilities = {}
            for c in self.classes:
                probabilities[c] = math.log(self.class_probabilities[c])
                for word in x.split():
                    if word in self.word_probabilities:
                        probabilities[c] += math.log(self.word_probabilities[word][c])
            result.append(max(probabilities, key=probabilities.get))
        return result

上述代码实现了一个朴素贝叶斯分类器，其中`fit`方法用于训练模型，`predict`方法用于预测新的数据。