1.导入数据
from sklearn.decomposition import PCA
from sklearn.ensemble import RandomForestClassifier as RFC
from sklearn.model_selection import cross_val_score
import matplotlib.pyplot as plt
import pandas as pd
import numpy as np
data=pd.read_csv(r"F:\sklearn文档\digit recognizor.csv")
x=data.iloc[:,1:]
y=data.iloc[:,0]
x.shape #(42000, 784)
(42000, 784)
2.画累计方差贡献率曲线,找最佳降维后维度范围
pca_line=PCA().fit(x)
plt.figure(figsize=[20,5])
plt.plot(np.cumsum(pca_line.explained_variance_ratio_))
plt.xlabel("number of components after dimension reduction")
plt.ylabel("cumulative explained variance")
plt.show()
3.画降维后维度的学习曲线,继续缩小最佳维度的范围
score=[]
for i in range(1,101,10):
x_dr=PCA(i).fit_transform(x)
once=cross_val_score(RFC(n_estimators=10,random_state=0),x_dr,y,cv=5).mean()
score.append(once)
plt.figure(figsize=[20,5])
plt.plot(range(1,101,10),score)
plt.show()
4.在[10,25]找出最佳维度
score1=[]
for i in range(10,25):
x_dr=PCA(i).fit_transform(x)
once=cross_val_score(RFC(n_estimators=10,random_state=0),x_dr,y,cv=5).mean()
score1.append(once)
plt.figure(figsize=[20,5])
plt.plot(range(10,25),score1)
plt.show()
由图可以看出,当维度去22时,模型表现效果达到最佳
x_dr=PCA(22).fit_transform(x)
cross_val_score(RFC(n_estimators=10,random_state=0),x_dr,y,cv=5).mean()
#改变随机森林n_estimators的参数值
cross_val_score(RFC(n_estimators=100,random_state=0),x_dr,y,cv=5).mean()
#使用KNN评估模型
from sklearn.neighbors import KNeighborsClassifier as KNN
cross_val_score(KNN(),x_dr,y,cv=5).mean()
--------------------------------------------------------------------------
随机森林(n_estimators=10) 0.9171190476190476
随机森林(n_estimators=100) 0.9447857142857142
KNN 0.9687619047619048
5.KNN画学习曲线
score2=[]
for i in range(10):
x_dr=PCA(22).fit_transform(x)
once=cross_val_score(KNN(i+1),x_dr,y,cv=5).mean()
score2.append(once)
plt.figure(figsize=[20,5])
plt.plot(range(10),score2)
plt.show()
当KNN参数值取2时,模型表现效果超过96%
cross_val_score(KNN(2),x_dr,y,cv=5).mean()
0.9689285714285714
在使用PCA后,原本784列的特征被我们缩减到22列之后,用KNN跑出了良好的效果,后续也可在使用PCA的基础上调整参数或改模型等方法提高模型的效果。
完!