Imagine a model is performing well on training data but poor... | Interview Question
IntermediateSITUATIONALTEXT
Imagine a model is performing well on training data but poorly on test data. How would you diagnose the issue and improve it?
ML/AI fresher
General
Sample Answer
This is based on the screening round and intermediate difficulty, which should test practical model troubleshooting rather than advanced specialization.
Related Keywords
What signs would tell you it is overfitting versus data leakage?Which fixes would you try first and why?
Tips for Answering
Follow these guidelines for a strong response
Use the STAR method (Situation, Task, Action, Result) for behavioral questions.
Provide specific examples from your experience with measurable results.
Keep your answer concise — aim for 1-2 minutes per response.
Practice out loud to improve delivery and identify gaps in your answer.