Python机器学习算法之K近邻算法

1.数据预处理

a.标准化

   

b.归一化

2.计算欧式距离

3.

<bound method NDFrame.head of       accommodates  bedrooms  bathrooms  beds    price  minimum_nights  \
0                4       1.0        1.0   2.0  $160.00               1   
1                6       3.0        3.0   3.0  $350.00               2   
2                1       1.0        2.0   1.0   $50.00               2   
3                2       1.0        1.0   1.0   $95.00               1   
4                4       1.0        1.0   1.0   $50.00               7   
5                4       2.0        1.0   4.0   $99.00               1   
6                4       2.0        2.0   2.0  $100.00               3   
7                2       1.0        1.0   1.0  $100.00               1   
8                2       1.0        1.5   1.0   $38.00               2   
9                2       1.0        NaN   1.0   $71.00               2   
10               4       2.0        1.5   2.0   $97.00               4   
11               1       1.0        1.0   1.0   $55.00               3   
12               2       1.0        1.0   1.0   $50.00               2   
13               2       0.0        1.0   1.0   $99.00               7   
14               2       1.0        NaN   1.0   $60.00               2   
15               2       1.0        1.0   1.0   $52.00               1   
16               1       1.0        1.0   1.0   $23.00               1   
17               2       1.0        1.0   2.0  $200.00               1   

3.算法实现

import pandas as pd
from sklearn.preprocessing import StandardScaler
from scipy.spatial import distance

dc_listings = pd.read_csv('listings.csv')
features = ['accommodates','bedrooms','bathrooms','beds','price','minimum_nights','maximum_nights','number_of_reviews']
dc_listings = dc_listings[features]  
dc_listings['price'] = dc_listings.price.str.replace("\$|,",'').astype(float)  
# 处理缺失值
dc_listings = dc_listings.dropna()
dc_listings[features] = StandardScaler().fit_transform(dc_listings[features])　　  #数据归一化
normalized_listings = dc_listings
# print(dc_listings.shape)
# normalized_listings.head()
# 分割训练集、测试集
norm_train_df = normalized_listings.copy().iloc[0:2792]
norm_test_df = normalized_listings.copy().iloc[2792:]
# print(norm_train_df.shape)
# print(norm_test_df.shape)
# 计算两个样本的欧氏距离
first_listing = normalized_listings.iloc[0][['accommodates','bathrooms']]
fifth_listing = normalized_listings.iloc[20][['accommodates','bathrooms']]
first_fifth_distance = distance.euclidean(first_listing,fifth_listing)

def predict_price_multivariate(new_listing_value,feature_columns):
temp_df = norm_train_df
temp_df['distance'] = distance.cdist(temp_df[feature_columns],[new_listing_value[feature_columns]]) 　　 #计算测试集合中某一个和训练集的欧式距离（多个属性）
temp_df = temp_df.sort_values('distance')
knn_5 = temp_df.price.iloc[:5]　　#取前5个最近邻，计算平均价格（预测）
predicted_price = knn_5.mean()
return(predicted_price)

cols = ['accommodates','bathrooms']
norm_test_df['predicted_price'] = norm_test_df[cols].apply(predict_price_multivariate,feature_columns=cols,axis=1)
norm_test_df['squared_error'] = (norm_test_df['predicted_price']-norm_test_df['price'])**2　　#预测价格和真实价格对比
mse = norm_test_df['squared_error'].mean()
rmse = mse**(1/2)
print(rmse)

# 利用sklearn来实现KNN
from sklearn.neighbors import KNeighborsRegressor
from sklearn.metrics import mean_squared_error
cols = ['accommodates','bathrooms']
knn = KNeighborsRegressor()
knn.fit(norm_train_df[cols],norm_train_df['price'])
two_features_predictions = knn.predict(norm_train_df[cols])

two_features_mse = mean_squared_error(norm_train_df['price'],two_features_predictions)
two_features_rmse = two_features_mse**(1/2)
print(two_features_rmse)