KNN决策树暴力求解

import numpy as np
import operator


# 数据集
def DataSet():
    # [打斗次数，kiss次数]
    # 矩阵9*2
    group = np.array([[3, 104], [2, 100], [1, 81], [101, 10], [99, 5], [98, 2], [10, 10], [80, 80], [20, 20]])
    # print(group)
    labels = ['love', 'love', 'love', 'move', 'move', 'move', 'love', 'love', 'move']
    return group, labels


def KNN(in_x, x_labels, y_labels, k):
    # 读取矩阵长度，shape[0]读取行数,shape[1]读取列数
    x_labels_size = x_labels.shape[0]
    # #np.tile(array1,(y轴方向复制次数，x轴方向复制次数))
    # 将待预测数据变为9*1的数组
    # （xi-xj）^2
    distances = (np.tile(in_x, (x_labels_size, 1)) - x_labels) ** 2
    # print(distances)
    # 每一行向量求和
    ad_distances = distances.sum(axis=1)
    # print(ad_distances)
    # 开方后得欧氏距离
    sq_distances = ad_distances ** 0.5
    # 对欧氏距离进行排序返回索引
    ed_distances = sq_distances.argsort()
    # 存储k个邻近点
    classdict = {}
    # 得到最近的k个数据的标签并对标签进行计数
    for i in range(k):
        # 得到对应的标签
        votel_label = y_labels[ed_distances[i]]
        # 对最近的k个数据的标签进行计数，准备多数表决
        #get(key,m) 反悔key的值，若无则返回m
        classdict[votel_label] = classdict.get(votel_label, 0) + 1
    # 多数表决
    sort_classdict = sorted(classdict.items(), key=operator.itemgetter(1), reverse=True)
    return sort_classdict[0][0]


if __name__ == '__main__':
    group, lables = DataSet()
    # x=input("请输入要预测的数据")
    test_x = [90, 90]
    print('输入数据所对应的类别是：{}'.format(KNN(test_x, group, lables, 3)))