Skip to content
InterviewEra
Loading account navigation
InterviewEra

InterviewEra is an AI-powered mock interview platform with adaptive follow-ups, resume-aware scoring, and structured interview preparation for campus placements and early-career hiring.

Start Mock Interview

Product

  • How It Works
  • For Teams
  • Start Mock Interview
  • Campus Placements
  • Campus Workspace
  • Help Center

Tools

  • Interview Question Generator
  • ATS Resume Checker
  • STAR Answer Builder

Resources

  • Interview Questions
  • All Resources
  • Blog
  • Community Hub
  • DSA Topic Map
  • Placement Guide
  • STAR Guide

Company

  • What is InterviewEra
  • About Us
  • Pricing
  • Contact
  • Privacy Policy
  • Terms & Conditions
  • Refund Policy

© 2026 InterviewEra. All rights reserved.

Privacy Policy|Terms & Conditions|Refund Policy
|Ranchi, Jharkhand, India
Interview Questions›Topics›Machine Learning

Data & ML

Machine Learning Interview Questions 2026

Machine Learning interview questions on supervised/unsupervised learning, model evaluation, regularisation, and algorithm trade-offs.

Data & ML

Commonly asked at: Commonly asked at data-heavy product companies such as Flipkart, Swiggy, and Amazon (for ML/data roles) — general pattern, not company-verified.

Company names below are mentioned only to indicate the general type/level of interview these questions are common at, based on widely known industry patterns. This content is not affiliated with, endorsed by, or sourced from any confidential material of the named companies. All trademarks belong to their respective owners.

Machine Learning Interview Questions

Placement-oriented · Updated 2026
  1. 01

    Supervised vs unsupervised learning — what's the difference?

    TechnicalEasy

    Tip: Give one example of each — house price prediction vs customer segmentation — it's the standard way to answer this.

    Spoken answer

    Supervised learning trains on labeled data, where each input has a known correct output — like predicting house prices from past sales with known prices. Unsupervised learning works with unlabeled data and finds patterns on its own, like clustering customers into segments without predefined categories.

    Point-wise answer

    • Supervised: labeled data, known correct outputs
    • Unsupervised: unlabeled data, finds patterns/structure
    • Examples: regression/classification vs clustering
  2. 02

    What is overfitting, and how do you prevent it?

    TechnicalEasy

    Tip: List at least 2-3 prevention techniques — a one-line answer feels incomplete here.

    Spoken answer

    Overfitting is when a model learns the training data too closely, including its noise, and performs poorly on new data. I'd prevent it with more training data, simplifying the model, adding regularization, using dropout in neural nets, and validating with cross-validation.

    Point-wise answer

    • Model memorizes training data instead of generalizing
    • Fixes: more data, regularization (L1/L2), dropout, simpler model
    • Detect via cross-validation / train-vs-test gap
  3. 03

    What is the bias-variance tradeoff?

    TechnicalEasy

    Tip: Define bias and variance separately first, then state the tradeoff — don't jump straight to "tradeoff."

    Spoken answer

    Bias is error from a model being too simple to capture the pattern — underfitting. Variance is error from being too sensitive to the training data's fluctuations — overfitting. Reducing one tends to increase the other, so the goal is finding the right balance for lowest total error on unseen data.

    Point-wise answer

    • Bias: underfitting, model too simple
    • Variance: overfitting, model too sensitive to training data
    • Goal: balance both to minimize error on new data
  4. 04

    Classification vs regression — what's the difference?

    TechnicalMedium

    Tip: Give one example each — spam detection vs price prediction.

    Spoken answer

    Classification predicts a discrete category, like spam vs. not spam. Regression predicts a continuous numeric value, like a house's price. The choice of algorithm and evaluation metric depends entirely on which type of problem you're solving.

    Point-wise answer

    • Classification: discrete labels/categories
    • Regression: continuous numeric values
    • Different algorithms/metrics apply to each
  5. 05

    What is a confusion matrix, and what do precision and recall mean?

    TechnicalMedium

    Tip: Explain precision and recall with the "false alarm vs missed case" framing — it's easier to remember than the formulas.

    Spoken answer

    A confusion matrix shows a classifier's predictions against actual outcomes — true/false positives and negatives. Precision measures how many predicted positives were actually correct, avoiding false alarms; recall measures how many actual positives were caught, avoiding missed cases — and there's usually a tradeoff between the two.

    Point-wise answer

    • Confusion matrix: TP, FP, TN, FN breakdown
    • Precision: correctness of positive predictions
    • Recall: coverage of actual positives
    • Tradeoff exists between the two
  6. 06

    What is gradient descent?

    TechnicalMedium

    Tip: Mention the learning rate explicitly — interviewers often follow up asking what happens if it's too high/low.

    Spoken answer

    Gradient descent is an optimization algorithm that minimizes a model's error by iteratively adjusting parameters in the direction that reduces the loss the most. The learning rate controls the step size — too high can overshoot, too low can take forever to converge.

    Point-wise answer

    • Iteratively minimizes the loss function
    • Adjusts parameters using the loss's gradient
    • Learning rate controls step size (tradeoff: speed vs stability)
  7. 07

    Decision tree vs random forest — what's the difference?

    TechnicalMedium

    Tip: Say random forest = many trees + averaging — that one-liner is the core of the answer.

    Spoken answer

    A decision tree splits data into branches to reach a prediction, but a single tree can easily overfit. A random forest builds many trees on random subsets of data and features, then averages or votes on their predictions — reducing overfitting and generally improving accuracy.

    Point-wise answer

    • Decision tree: single tree, prone to overfitting
    • Random forest: many trees + random subsets, averaged/voted
    • Reduces overfitting vs a single tree
  8. 08

    What is feature engineering, and why does it matter?

    TechnicalMedium

    Tip: Give one concrete example — extracting "day of week" or "BMI" — abstract explanations alone feel weak here.

    Spoken answer

    Feature engineering is creating, transforming, or selecting input variables to improve model performance — like extracting "day of week" from a timestamp, or combining height and weight into a BMI feature. Good feature engineering often improves accuracy more than switching to a fancier algorithm.

    Point-wise answer

    • Creating/transforming input variables for better signal
    • Example: timestamp → day-of-week, height+weight → BMI
    • Often more impactful than algorithm choice alone
  9. 09

    What is regularization (L1 vs L2)?

    TechnicalMedium

    Tip: Mention that L1 can zero-out features (feature selection) while L2 shrinks them — that distinction is commonly asked.

    Spoken answer

    Regularization adds a penalty term to the loss function to discourage overly complex models and reduce overfitting. L1 (Lasso) can shrink some coefficients all the way to zero, effectively doing feature selection, while L2 (Ridge) shrinks coefficients smoothly without zeroing them out.

    Point-wise answer

    • Adds penalty to loss function to reduce overfitting
    • L1 (Lasso): can zero out coefficients (feature selection)
    • L2 (Ridge): shrinks coefficients, doesn't zero them out
  10. 10

    ⭐ Scenario: A model performs great on training data but poorly in production. How would you debug this?

    SituationalHardSTAR

    Tip: This is scenario-based — walk through your diagnostic checklist step by step.

    Situation: A model showed strong accuracy during training/validation but underperformed once deployed.

    Task: I needed to find why the gap existed and close it without just retraining blindly.

    Action: I checked whether production data distribution had drifted from training data, verified there was no data leakage in the original training/validation split, and confirmed the exact same preprocessing pipeline was applied at inference time as during training.

    Result: I found a preprocessing mismatch — a feature was being computed slightly differently in production — and fixing that closed most of the performance gap.

Practice ML questions with your own resume

InterviewEra generates role-specific questions using your actual projects and skills. Get scored feedback on technical depth, clarity, and structure — free to start.

Start free mock interviewFree question generator

Roles that need ML

  • Python Developer questions
  • Data Analyst questions
  • Machine Learning Engineer questions
  • Data Scientist questions

Related data topics

  • Pandas questions

Practice tools

  • Interview question generator
  • ATS resume checker
  • STAR answer builder

Guides and resources

  • All interview questions
  • HR interview answer tips
  • STAR method with examples