Explain leakage, missing data, and common losses
Company: Adobe
Role: Machine Learning Engineer
Category: Machine Learning
Difficulty: medium
Interview Round: Technical Screen
Answer the following traditional ML questions:
1. **Data leakage**
- What is data leakage?
- Give 2–3 common examples.
- How do you prevent or fix it in practice?
2. **Missing data**
- What are common strategies to handle missing values?
- When might you drop rows/columns vs impute?
- How can missingness itself be informative?
3. **Linear vs logistic regression losses**
- What loss is commonly used for linear regression? For logistic regression?
- Compare **MSE vs MAE**: how do they differ, and when might you prefer one?
Quick Answer: This question evaluates a candidate's understanding of data leakage, strategies for handling missing data, and the differences between loss functions used in linear and logistic regression (including trade-offs between MSE and MAE), probing competency in data preprocessing, robustness, and model evaluation within Machine Learning.