Factors Involved in ML Algorithm Selection CT-AI Practice Questions (Page 1)

Questions 1–5

1. 1foundation · easy

   Which algorithm family is most appropriate for predicting a continuous value, such as house price?

   • ALogistic regression
   • BK-means clustering
   • CLinear regression
   • DDecision tree for classification
   Reveal Solution Discussion Select an answer first
2. 2expert · hard

   A data science team is building a model to predict loan default. They have a dataset with 100,000 records and 50 features. They have tried a single decision tree and a logistic regression model, but both have high bias (underfitting). They need to improve accuracy without overfitting. They have moderate computational resources. Which approach is most appropriate?

   • AUse gradient boosting with a low learning rate and early stopping
   • BUse a support vector machine with RBF kernel
   • CUse a deeper decision tree
   • DUse a random forest with more trees
   Reveal Solution Discussion Select an answer first
4. 3foundation · easy

   A data scientist is selecting an ML algorithm for a project with 10,000 rows and 500 features. Which factor is most directly relevant to this choice?

   • AThe number of developers on the team
   • BThe programming language used for the web front-end
   • CThe color scheme used in data visualizations
   • DThe number of features relative to the number of rows
   Reveal Solution Discussion Select an answer first
5. 4foundation · easy

   When is it most appropriate to choose an ensemble method over a single model?

   • AWhen training time must be minimized
   • BWhen interpretability is the top priority
   • CWhen the single model has high variance and you need to improve generalization
   • DWhen the dataset is very small and simple
   Reveal Solution Discussion Select an answer first
6. 5expert · hard

   A retail company wants to predict which products a customer will buy next (multi-class classification with 100 product categories). They have a dataset of 10 million transactions with sparse categorical features. The model must be retrained weekly on a standard cloud cluster. The business needs the model to be interpretable for marketing insights. Which algorithm is most suitable?

   • AMultinomial logistic regression
   • BSupport vector machine with one-vs-rest
   • CK-nearest neighbors
   • DDeep neural network with softmax output
   Reveal Solution Discussion Select an answer first