ML Theory Interview Questions [2026-2027]

Q1. What is the difference between AI, Data Science, ML, and DL? Ans: Artificial intelligence (AI) is the ability of a computer or a robot controlled by a computer to do...

Here's an update on a less common interview experience with Luma AI, they've been quite popular lately. Please give me some points! The screening round was with the hiring manager (HM). They asked abo

The interview for an ML generalist started with simple background questions, followed by some follow-up questions about model architecture/deployment. Then they tested me on NumPy array operations. I

This post was last edited by Anonymous on 2025-10-09 11:19 OA: This has been posted many times on the forum. One question was linear interpolation, another was pandas data processing + scikit-learn li

This post was last edited by an anonymous forum user on 2025-09-27 20:06. Five rounds: one coding round, one ML system design round, one ML model design round, and one behavioral questions (BQ) round.

The questions seem to be randomly selected from a question bank, divided into statistics, machine learning, and coding questions. The overall difficulty isn't high. I've included a few screenshots. Ne

This was my first online assessment (OA) since applying. I applied on October 6th and received the invitation on October 10th, completing it within three days. There were three questions. The first wa

I applied to this company last year and got an interview, but withdrew because I thought the process was too long. I reapplied this year and was scheduled for an interview the next day. The first roun

These questions were randomly selected from a question bank and are divided into statistics, machine learning, and coding problems. The overall difficulty is not high. I've included a few screenshots.

Hello, all in the following article I'm putting my interview experience at Dell Technologies. At first, it comes for another job profile but after giving online tests they...

Ericsson Global visited our campus this October. The whole process was virtual.Eligibility: 7 CGPA and above. CSE, IT, and ECE.There were 3 rounds in all for the hiring pr...

Even if we were to search around the world, it would be a truly difficult job to find someone like Divanshu. As a Computer Science Graduate of IIIT – Allahabad, he succes...

How do they make those Awesome Animation movies? How does it feel to work for a film like ‘How to Train your Dragon 2 ‘ or ‘Penguins of Madagascar’? This guy has an answer...

Anudeep Nekkanti embodies the old adage – there is no substitute for talent. The 21-year-old coder from Samalkot (a small town near Vishakhapatnam) has landed an offer fro...

Technical + Managerial Round:I interviewed with Honeywell for the position of Software Engineering Co-op. It was a 30 minutes scheduled round with two experienced engineer...

Lowe's India came for Campus Recruitment during placement season for the post of Data Scientist at IISc Bangalore. Following are the complete details about the interview p...

Position: Software Engineer Duration: July 2021 - PresentOverview: I began my journey at Sopra Steria, a leading global IT consulting and services company, as a fresher an...

There were 4 rounds including an online test.Round 1: This round had 50 MCQ and 2 coding questions. MCQ mainly composed of technical and aptitude questions. The technical ...

Round 1: Assignment.You will get an assignment form HR to solve. For my case it was extraction of data from doc format. The crux of clearing this round lies in labeling th...

Round 1: Shell was on campus for recruitment for the post of SDE. The eligible branches were CSE and IT only. The first round had 700 students shortlisted from both the br...

Round 1: Normal aptitude consisting of Math and Java ( Refer India Bix)40 were selected out of 300Round 2: Group Discussion4 groups of 10 each were formed and topic were o...

Written Test:Questions asked From fallowing concepts:1. Aptitude2. C aptitude Questions (very easy vectorians easily crack those)3. OS basic commandsQ1. How wil...

Shell Technology Centre Bangalore visited our college for recruiting Software Engineer. They conducted only two rounds: Online round and Personal Interview.Total 60 studen...

## Problem You are asked to build a CTR (click-through rate) prediction model for a content recommendation system. Walk through the full ML modeling process: **1. Problem framing** - Binary classification: will user click on item? (positive = click) - Training signal: implicit feedback (clicks), with heavy class imbalance (~1% CTR) **2. Feature engineering** - User features: historical CTR, session recency, demographics - Item features: category, age, historical CTR - Context features: device, time-of-day, position bias **3. Model options** - Logistic Regression (baseline, interpretable) - Gradient Boosted Trees (GBDT) for tabular features - Deep factorization machines or two-tower neural model for large sparse IDs **4. Evaluation** - Metrics: AUC-ROC, log-loss, calibration - Why accuracy is a poor metric at 1% CTR ## Follow-ups 1. How do you handle position bias in training data (items shown higher get more clicks)? 2. How do you evaluate the model offline before an A/B test? 3. The model's calibration drifts after 2 weeks — what causes this and how do you fix it?

## Problem You are a machine learning engineer inheriting a dataset used to train a churn prediction model. The dataset has columns: `user_id`, `age`, `tenure_days`, `monthly_spend`, `support_tickets`, `churned` (0/1). You run a data health check and find: - 12% of `age` values are missing. - `monthly_spend` has a right skew with outliers at 100x the median. - `support_tickets` has zero-inflation (80% of rows are 0). - The churn rate is 4% (class imbalance). **For each issue, state:** 1. Why it is a problem for model training. 2. At least two strategies to address it. 3. Which strategy you would recommend and why. ## Follow-ups 1. How would you impute `age` - mean, median, mode, or model-based? What are the tradeoffs? 2. For the spending outliers, compare log-transform vs. Winsorization vs. removing the rows. 3. What resampling techniques (SMOTE, undersampling) help with the 4% churn imbalance? What metric should you optimize instead of accuracy? 4. How would you set up automated data quality checks so these issues are caught before each training run?