Question $1$:

Why does data augmentation help improve a machine learning model$$s generalization performance?

a$.$ It reduces the need for model evaluation on test data

b$.$ It introduces random noise into the training data

c$.$ It artificially increases the training dataset$$s size and diversity

d$.$ It replaces the need for hyperparameter tuning

Explain your answer

Question $2$:

How does decreasing model complexity typically affect bias and variance in a machine learning model?

a$.$ Bias increases and variance decreases

b$.$ Bias decreases and variance increases

c$.$ Both bias and variance increase

d$.$ Both bias and variance decrease

Explain your answer

Question $3$:

Why do we perform hyperparameter tuning in machine learning?

a$.$ To adjust the model$$s internal parameters during testing

b$.$ To fine$-$tune external model settings for improved model performance

c$.$ To adjust the model$$s features for better accuracy

d$.$ To reduce the complexity of the training data

Explain your answer

Question $4$:

Why would you use the k$-$fold cross$-$validation method for validating your model performance over a regular train$-$validate$-$test data split?

a$.$ Reduce overfitting the data

b$.$ Reduce dependency on a single split that could be biased or unrepresentative

c$.$ Faster run time and less need for computational resources

d$.$ Eliminate the need for a validation dataset

Explain your answer