Machine Learning Interview Question

Question

If you are having 4GB RAM in your machine and you want to train your model on 15GB data set. What would be your approach?

in progress 0
pritam07 4 years 1 Answer 790 views Newbie 0

Answer ( 1 )

  1. For Neural network –
    1)Firstly load the complete data using Numpy Array. Numpy Array helps in mapping the dataset without completely applying it to memory.
    2)After that just pass the index to view the Numpy array.
    3)Have a small batch size.
    For SVM
    1)Divide datasets into small datasets.
    2)Use a partial fit method of SVM, it requires a subset of the complete data set.

Leave an answer

Browse
Browse