Loading Question...
Basic ML and stats questions were asked p test,t test, distributions(Poisson,normal etc.),central limit theorem,then boosting vs bagging, then went bit deeper into them individually starting with xgboost,adaboost , decision trees ( entropy,gini impurity) , then coming to deep learning, cnn( padding,striding etc.),lstm,transformers basic structure and difference, vanishing gradient, exploding gradient problem why do they happen etc. and finally asked a python related question ( he orally mentions the dataframe) and i had to just tell the approach
Round 2 was a classification task, dataset was shared first i was asked to interpret the dataset then basic feature engineering,eda etc. and training a simple logistic regression was enough ( looking upto official pandas,numpy docs were allowed)
Round 3 was about cv discussion, projects and again basic ml and dl questions
Last updated Oct 8, 2026
Hr round, didnt make it
Tips: There is a repo by a bhu senior who works at infoedge which was pretty helpful also for first and third round, just revising the oa questions and concepts would be helpful they asked questions around those only. For the second round, the model training part i would say practice once on say titanic dataset and by practice i mean even if you arenot that comfortable with it you could memorise basic eda and feature engineering templates and practice em on those kaggle datasets because you will have to narrate what you are doing as well.