SAP interview Questions | hyperparameters

Question

For tuning hyperparameters of your machine learning model, what will be the ideal seed?

in progress 0
Dhruv2301 4 years 2 Answers 738 views Great Grand Master -1

Answers ( 2 )

  1. swap007
    0
    Reply

    Seed is ideally used so that when you re-run your models , the same set of observations
    go in the train and test set and you are able to replicate the results. So, you can set any seed but
    stick to it throughout the process.

  2. Priyabrata017
    0
    Reply

    We can take any seed. There is no hard and fast rule that for hyper paramneter tuning, a particular seed has to be chosen. But we have to make sure the seed is same for all the models.

Leave an answer

Browse
Browse