Section 3, Unit 1

AI & Machine learning
1

Hi everyone just to clarify: the point made at the end of the video is whether or not to shuffle the dataset before. splitting it? 



Perhaps consider displaying the shuffle option in the train_test_split API for clarity. 



Thanks,

Alex

2

Hello Alex,



The last point in the video in section 3 unit 1 is about class weight. This parameter is set along with the weights of various classes especially when there is a class imbalance in the dataset. This is used when calculating the splitting criterion and generally we pass the inverse frequency of each class as weight.