K近邻算法_分类鸢尾花数据集

import numpy as np
import pandas as pd 
from sklearn.datasets import load_iris
from sklearn.model_selection import train_test_split
from sklearn.metrics import accuracy_score

1.数据预处理

iris = load_iris()
df = pd.DataFrame(data=iris.data, columns=iris.feature_names)
# 分类标签数据
df['class'] = iris.target
# 数值转为文字
df['class'] = df['class'].map({0: iris.target_names[0], 1: iris.target_names[1], 2: iris.target_names[2]})
df.head()

x = iris.data
y = iris.target.reshape(-1, 1)
print("x shape: " , x.shape)
print("y shape: ", y.shape)

x shape:  (150, 4)
y shape:  (150, 1)

# 划分训练集和测试集
x_train, x_test, y_train, y_test = train_test_split(x, 
        y,test_size=0.3, random_state=42, stratify=y)

2. 模型实现

# L1 距离
def l1_distance(a, b):
    return np.sum(np.abs(a - b), axis = 1)

#L2 距离
def l2_distance(a, b):
    return np.sqrt(np.sum((a - b)**2, axis = 1))

# K近邻模型
class KnnModel(object):
    def __init__(self, k_neighbors = 1, distance_func = l1_distance):
        self.k_neighbors = k_neighbors;
        self.distance_func = distance_func
    
    #不需要训练,只是预测时用于计算预测点的距离
    def fit(self, x, y):
        self.x_train = x
        self.y_train = y
    
    def predict(self, test):
        y_predict = np.zeros((test.shape[0],1), dtype=self.y_train.dtype)
        for i, x_test in enumerate(test):
            # 计算 测试点和训练集的距离
            distances = self.distance_func(self.x_train, x_test)
            # 按照距离大小排序，取出索引
            sort_index = np.argsort(distances)
            # 取出前 k 个值
            neighbors_predict = self.y_train[sort_index[:self.k_neighbors]].ravel()
            # 取出前 k 个值里面出现最多的数
            y_predict[i] = np.argmax(np.bincount(neighbors_predict))
        return y_predict
            
        

3.测试

knn = KnnModel(k_neighbors = 9)
knn.fit(x_train, y_train);

result_list = []
for df in [1, 2]:
    knn.distance_func = l1_distance if pd == 1 else l2_distance
    
    for k in range(1, 20 , 2):
        knn.k_neighbors = k
        y_predict = knn.predict(x_test)
        acc = accuracy_score(y_test, y_predict) * 100
        result_list.append([k, 'l1_dist' if df == 1 else 'l2_dist', acc])
        
result_df = pd.DataFrame(result_list, columns=['k', '距离函数', '准确率'])
print(result_df)

     k     距离函数        准确率
0    1  l1_dist  93.333333
1    3  l1_dist  95.555556
2    5  l1_dist  97.777778
3    7  l1_dist  95.555556
4    9  l1_dist  95.555556
5   11  l1_dist  93.333333
6   13  l1_dist  93.333333
7   15  l1_dist  95.555556
8   17  l1_dist  95.555556
9   19  l1_dist  95.555556
10   1  l2_dist  93.333333
11   3  l2_dist  95.555556
12   5  l2_dist  97.777778
13   7  l2_dist  95.555556
14   9  l2_dist  95.555556
15  11  l2_dist  93.333333
16  13  l2_dist  93.333333
17  15  l2_dist  95.555556
18  17  l2_dist  95.555556
19  19  l2_dist  95.555556