The difference between stratifiedkfold (classification) and kfold (regression)

One 、 StratifiedKFlod And KFlod The main difference

StratifiedKFlod： Stratified sampling , The comparison of samples in each category in the training set and the test set is the same as that in the original data ;（ Classification problem ）
KFlod： Stratified sampling , Divide the data into training set and test set , Whether the data of each category in the training set and the test set are the same is not considered ;（ The return question ）

from sklearn.model_selection import KFold,StratifiedKFold

KFold（n_split, shuffle, random_state）
Parameters ：
n_splits： It means dividing the data into several equal parts
shuffle： In each division , Whether to shuffle
if False, The effect is equivalent to random_state Integers ( With zero ), The result of each division is the same
if True, The result of each division is different , Indicates that the cards have been shuffled , Random sampling
random_state： Random seed number , When the set value ( It's usually 0) It is convenient to adjust parameters after , Because the data set generated each time is the same

stratifiedKFold（n_split, shuffle, random_state）
Parameters ：
n_splits： It means dividing the data into several equal parts
shuffle： In each division , Whether to shuffle
if False, The effect is equivalent to random_state Integers ( With zero ), The result of each division is the same
if True, The result of each division is different , Indicates that the cards have been shuffled , Random sampling
random_state： Random seed number , When the set value ( It's usually 0) It is convenient to adjust parameters after , Because the data set generated each time is the same

Two 、 StratifiedKFlod And KFlod Different cases

import numpy as np 
from sklearn.model_selection import KFold,StratifiedKFold
X=np.array([
    [1,2,3,4],
    [11,12,13,14],
    [21,22,23,24],
    [31,32,33,34],
    [41,42,43,44],
    [51,52,53,54],
    [61,62,63,64],
    [71,72,73,74]
])
y=np.array([1,1,0,0,1,1,0,0])

KFold = KFold(n_splits=4,shuffle=True,random_state=2021)
StratifiedKFold = StratifiedKFold(n_splits=4,shuffle=True,random_state=2021)
print('---------------------KFlod---------------------------')
for train, test in KFold.split(X,y):
    print('Train: %s | test: %s' % (train, test))
    print(' Training set label type ： %s' % y[train])
    print(' Test set label type ： %s' % y[test])
print('----------------StratifiedKFold----------------------')
for train, test in StratifiedKFold.split(X,y):
    print('Train: %s | test: %s' % (train, test))
    print(' Training set label type ： %s' % y[train])
    print(' Test set label type ： %s' % y[test])

#  The input result is as follows 
''' ---------------------KFlod--------------------------- Train: [0 1 2 4 5 6] | test: [3 7]  Training set label type ： [1 1 0 1 1 0]  Test set label type ： [0 0] Train: [0 1 3 4 5 7] | test: [2 6]  Training set label type ： [1 1 0 1 1 0]  Test set label type ： [0 0] Train: [2 3 4 5 6 7] | test: [0 1]  Training set label type ： [0 0 1 1 0 0]  Test set label type ： [1 1] Train: [0 1 2 3 6 7] | test: [4 5]  Training set label type ： [1 1 0 0 0 0]  Test set label type ： [1 1] ----------------StratifiedKFold---------------------- Train: [0 1 2 3 4 6] | test: [5 7]  Training set label type ： [1 1 0 0 1 0]  Test set label type ： [1 0] Train: [0 1 2 3 5 7] | test: [4 6]  Training set label type ： [1 1 0 0 1 0]  Test set label type ： [1 0] Train: [0 3 4 5 6 7] | test: [1 2]  Training set label type ： [1 0 1 1 0 0]  Test set label type ： [1 0] Train: [1 2 4 5 6 7] | test: [0 3]  Training set label type ： [1 0 1 1 0 0]  Test set label type ： [1 0] '''

summary ：
KFlod It is applicable to user regression type data division
stratifiedKFlod Applicable to classification data division

Reference resources ：
https://blog.csdn.net/qq_34107425/article/details/105548800
https://blog.csdn.net/wqh_jingsong/article/details/77896449