AI and machine learning interview questions

Pass the whole Pipeline to GridSearchCV or RandomizedSearchCV.

Backpropagation applies the chain rule to compute the gradient of the loss with respect to every weight.

In deep sigmoid or tanh stacks, gradients shrink as they go backward, so early layers barely learn.

An epoch is one full pass over the training set.

Elastic net mixes L1 and L2.

For vanilla SGD they match: both add a term that pulls weights toward zero.

Stopping before the optimizer fully minimizes train loss limits effective capacity.

Use stratified k-fold for classification so each fold keeps the class mix.

GroupKFold keeps all rows from the same group, such as a user or a hospital, on one side of the split.

ShuffleSplit draws repeated random train and val cuts, which is flexible when you want a fixed train size.

On imbalanced data, predicting the majority class always looks accurate but useless.

Scaling puts numeric features on similar ranges so distance-based and gradient methods behave well.

It summarizes how well scores rank positives above negatives across thresholds.

k-NN predicts from the labels of the k closest training points.

Add trees that correct previous residuals step by step.

Encoding categories with target means can leak if done before split.

Resample training data, class weights, or better metrics.

Estimating generalization by rotating folds.

Without non-linearities, stacked layers collapse to one linear map.

When false positives are expensive — like blocking legit users.