ML Workflow CT-AI Practice Questions (Page 2)

Questions 6–10

1. 6application · medium

   A team is training a neural network to recognize images. They notice the training loss decreases but the validation loss increases after a few epochs. What is the most appropriate action to address this issue?

   • AIncrease the learning rate to speed up convergence
   • BIncrease the number of training epochs to allow the model to converge further
   • CStop training early and use the model from the epoch with the lowest validation loss
   • DRemove the validation set and train only on the training data
   Reveal Solution Discussion Select an answer first
2. 7foundation · easy

   In an ML workflow, what is the purpose of the model evaluation step?

   • ATo collect additional data from various sources for training
   • BTo deploy the model into a production environment
   • CTo clean and transform raw data into a suitable format
   • DTo assess the model's performance on unseen data before deployment
   Reveal Solution Discussion Select an answer first
4. 8application · medium

   A data scientist has trained a binary classifier and obtained a confusion matrix. The model has high precision but low recall. The business context is that false negatives are very costly (e.g., missing a disease). What should the scientist do?

   • ARaise the classification threshold to increase precision
   • BKeep the threshold unchanged to maintain current performance
   • CUse a different evaluation metric, such as accuracy
   • DLower the classification threshold to increase recall
   Reveal Solution Discussion Select an answer first
5. 9application · medium

   A data scientist is working with a dataset that has a feature with a highly skewed distribution. The scientist plans to use a model that assumes normally distributed features. What is the most appropriate transformation?

   • AStandardize the feature by subtracting the mean and dividing by the standard deviation
   • BLeave the feature as is, as most models are robust to skewness
   • CRemove the feature from the dataset
   • DApply a logarithmic transformation to reduce skewness
   Reveal Solution Discussion Select an answer first
6. 10foundation · easy

   Which of the following is a key input to the model training process?

   • ATraining data with known outcomes (labels)
   • BA deployed model in a production environment
   • CA monitoring dashboard for performance tracking
   • DA test set used for final evaluation
   Reveal Solution Discussion Select an answer first