1. Python Coding - Pandas, List Operations, DSA 2. Statistics & Probability - CLT, Bayesian, Conditional, etc. 3. ML Questions - Logistics Regression, Evaluation Metrics, Decision Trees, Optimisation 4. Analytics & Probability Puzzles 5. How do you deal with difficult ppl in teams, etc.
why should you want to be a part of EXL Family
How would you convince upper management to provide you with funding for a project. What if they disagree with your approach, how would you go about convincing them?
Project based, Transformer, BERT architecture based
why accuracy is not a good metric? in what cases recall is a better metric? what is p-value? how is lgbm different from gbm, xgboost?
Should you be concerned with capping and flooring variables for a random forest model?
All about dealing with conflict, because they have a culture of conflict.
1. How is word transformers different from Sentence transformers. 2. Scenario based questions. 3. When to use F1 Score, Precision, Recall Score 4. Clustering before train test split or after train test split. 5. I was surprised on this question. Do SMOTE increase your test data sample size as well 6. Scaling - MinMax, Standard and Robust Scaler 7. When to use class weights and oversampling/Undersampling ( No one can get it right at the first time). I wont recommend for someone who has good experience in Data Science. In my decade of taking/giving data Science interviews, this has to be rated as my worst experience.
Forests, boosting, projects from CV.
How have you managed high achievers vs low performers in your team?
Viewing 101 - 110 interview questions