In Round 1, the technical interview questions include a self-introduction, explaining an ML project done at work and one done personally, NLP experience, word representation in NLP, hypothesis testing experience, explaining a non-deep learning project, a quiz on basic ML algorithms, statistical hypothesis testing, and Pandas, and finally implementing IoU in Python.
In Round 2, the technical interview questions include a self-introduction, explaining a project at work that was deployed in production, box plot interpretation and skew, CNNs, explaining an ML project with tabular data, hyperparameters in XGBoost and RF, what happens when increasing depth and n_estimators in RF when to use tree-based models, explaining backpropagation in NNs, regularization (Ridge vs Lasso) and how it helps overcome overfitting, and finally a walkthrough of one of my Kaggle projects.
Round 3 is a techno-managerial interview that covers interests, motivation, and an end-to-end scenario for building an ML model.
