双12购车预测推荐

import pandas as pd
from sklearn.metrics.pairwise import cosine_similarity

# 假设我们有一个用户-车辆评分矩阵
data = {
    'user_id': [1, 1, 2, 2, 3, 3],
    'car_id': [101, 102, 101, 103, 102, 104],
    'rating': [5, 3, 4, 1, 5, 2]
}
df = pd.DataFrame(data)

# 创建用户-车辆评分矩阵
user_car_matrix = df.pivot(index='user_id', columns='car_id', values='rating').fillna(0)

# 计算余弦相似度
similarity_matrix = cosine_similarity(user_car_matrix.T)

# 推荐函数
def recommend_cars(user_id, similarity_matrix, user_car_matrix, top_n=3):
    user_index = user_id - 1  # 索引从0开始
    similar_users = similarity_matrix[user_index].argsort()[::-1][1:]  # 排除自身
    recommended_cars = set()
    
    for similar_user in similar_users:
        similar_user_cars = user_car_matrix.iloc[similar_user].dropna().index
        recommended_cars.update(similar_user_cars)
        if len(recommended_cars) >= top_n:
            break
    
    return list(recommended_cars)[:top_n]

# 示例调用
recommended_cars = recommend_cars(user_id=1, similarity_matrix=similarity_matrix, user_car_matrix=user_car_matrix)
print(f"为用户{user_id}推荐的车辆ID: {recommended_cars}")