# KNN算法的封装
import numpy as np
from math import sqrt
from collections import Counter
# 定义分类knn函数
def KNN_classify(k,X_tarin,y_train,x):
# 断言判断用户参数传入是否正确
assert 1<=k<=X_tarin.shape[0],"k must be valid"
assert X_tarin.shape[0]==y_train.shape[0],\
"the size of X_tarin must equal to the size of y_train"
assert X_tarin.shape[1]==x.shape[0],\
"the feature number of x must equal to X_train"
# 求得最近距离
distance=[sqrt(np.sum((x_train-x)**2)) for x_train in X_tarin]
# 对所求最近距离点进行字典排序
nearest=np.argsort(distance)
topK_y=[y_train[i] for i in nearest[:k]]
# 找出出现次数最多的
votes=Counter(topK_y)
# 返回样本标签
return votes.most_common(1)[0][0]
if __name__ == '__main__':
# 样本数据
X_tarin=[[100,200],[110,300],[120,400],[130,500],[140,600],[9,10],[9,8],[7,8],[8,9],[8,7]]
X_data=np.array(X_tarin)
# 样本标签
y_tarin=[0,0,0,0,0,1,1,1,1,1]
y_data=np.array(y_tarin)
# 测试数据
x=([3,4])
x=np.array(x)
# 进行预测
predict_y=KNN_classify(5,X_data,y_data,x)
print(predict_y)
KNN算法的封装
最新推荐文章于 2023-04-24 09:20:03 发布