Hw5 q3.2


#1

When we need to train a binary classifier between the training and test sets, does the lambda pertain to (1) how to bias our training set or (2) how to choose rows from the training vs. test set?

I’m assuming we need to create a new dataset that has some rows from the test dataset and some from the training dataset in order to train our classifier, but I’m just wondering if lambda relates to how to choose rows from the training vs. test sets or if it still applies to how to bias our training data.


#2

Can you explain the meaning of the note for question5.3?
Note - you need to weigh the data.


#3

See here