西瓜数据集分类：Python 代码实现朴素贝叶斯算法

首先，我们需要计算数据集中每个特征的类别数量和类别比例。下面的代码实现了这一步骤：

class_counts = {}
for data in dataset:
    label = data[-1]
    if label not in class_counts:
        class_counts[label] = 0
    class_counts[label] += 1

class_ratios = {}
for label, count in class_counts.items():
    class_ratios[label] = count / len(dataset)

接下来，我们需要计算每个特征的每个类别的数量和比例。下面的代码实现了这一步骤：

feature_counts = {}
feature_ratios = {}

for feature in features:
    feature_counts[feature] = {}
    feature_ratios[feature] = {}
    for data in dataset:
        label = data[-1]
        value = data[features.index(feature)]
        if value not in feature_counts[feature]:
            feature_counts[feature][value] = {}
            feature_ratios[feature][value] = {}
        if label not in feature_counts[feature][value]:
            feature_counts[feature][value][label] = 0
            feature_ratios[feature][value][label] = 0
        feature_counts[feature][value][label] += 1

for feature in features:
    for value in feature_counts[feature]:
        total_count = sum(feature_counts[feature][value].values())
        for label in feature_counts[feature][value]:
            feature_ratios[feature][value][label] = feature_counts[feature][value][label] / total_count

现在我们可以使用这些统计信息来进行分类。我们可以使用朴素贝叶斯算法，根据先验概率和条件概率来计算后验概率，并选择后验概率最大的类别作为预测结果。下面的代码实现了这一步骤：

def predict(data):
    probabilities = {}
    for label in class_counts:
        probabilities[label] = class_ratios[label]
        for feature in features:
            value = data[features.index(feature)]
            probabilities[label] *= feature_ratios[feature][value][label]
    return max(probabilities, key=probabilities.get)

data = ['浅白', '稍蜷', '浊响', '模糊', '平坦', '硬滑']
prediction = predict(data)
print(prediction)

运行代码，可以得到预测结果为'否'。