Hands-On Exercise: Build a Defect Prediction System CT-AI Practice Questions (Page 5)

Questions 21–25

1. 21application · medium

   A team has collected a defect dataset where only 5% of modules have defects. They plan to train a classifier to predict defective modules. Which preprocessing step is most critical to avoid a model that simply predicts 'no defect' for everything?

   • AApply a resampling technique such as SMOTE to balance the classes.
   • BNormalize all features to a range of 0 to 1.
   • CEncode categorical features using one-hot encoding.
   • DRemove all modules with missing values from the dataset.
   Reveal Solution Discussion Select an answer first
2. 22application · medium

   A defect prediction model outputs a probability score for each module. The QA team has limited testing resources and wants to focus on the modules most likely to contain defects. How should they use the model's output?

   • AAverage the predictions with random selection to avoid bias.
   • BSort modules by predicted probability and test the highest-scoring ones first.
   • CTest modules in alphabetical order to ensure fairness.
   • DUse the model's predictions to skip testing entirely for low-scoring modules.
   Reveal Solution Discussion Select an answer first
4. 23application · medium

   A defect prediction model is evaluated on a test set. The team wants to minimize the risk of missing defective modules that could cause production failures. Which evaluation metric should they prioritize?

   • APrecision
   • BF1-score
   • CAccuracy
   • DRecall (sensitivity)
   Reveal Solution Discussion Select an answer first
5. 24foundation · easy

   What is the primary purpose of a defect prediction system in software testing?

   • ATo measure the number of defects found during testing
   • BTo replace manual testing with automated test execution
   • CTo identify areas of the code that are likely to contain defects so that testing efforts can be focused there
   • DTo automatically fix defects before they are reported by users
   Reveal Solution Discussion Select an answer first
6. 25foundation · easy

   In a typical defect prediction workflow, what is the correct sequence of steps after data collection?

   • AModel evaluation, data preprocessing, model training, deployment
   • BData preprocessing, model evaluation, model training, deployment
   • CModel training, data preprocessing, model evaluation, deployment
   • DData preprocessing, model training, model evaluation, deployment
   Reveal Solution Discussion Select an answer first