Interview questions

Compare trees, RF, and gradient boosting

General practice

Compare two rare-event detection models statistically

Companies: Waymo

Hypothesis Testing · Model Evaluation

Compare XGBoost and LightGBM

Companies: Turo

Compute and plot a precision–recall curve

Companies: Microsoft

ML Coding · Model Evaluation

Compute EL and RWA from loan data

Companies: Wells Fargo

Compute gambler’s ruin probabilities and hitting times

Companies: Imc

Compute Random Walk Hitting and Return Probabilities

Companies: Cfm

Compute Recall and FPR from Confusion Matrices to Select Classifiers

Companies: Pinterest

Model Evaluation

Compute value of card guessing game

Companies: Gts

Connect Bias-Variance Trade-offs to Bagging and Boosting

Companies: C3 Ai

Construct a Churn-Prediction Pipeline Using Scikit-Learn

Companies: Apple

Feature Engineering · Supervised Learning

Contrast Lasso vs Ridge trade‑offs

Companies: Instacart

ML Fundamentals · Regression & Statistical Modeling

Contrast LSTM and Transformer for long sequences

Companies: TikTok

Core algorithms inside a demand-side platform: bidding, prediction, pacing and fraud

Companies: Rakuten

Ads Systems · Technical Tradeoffs for PMs

Core Components of an LLM-Based AI Agent and How They Interact

Companies: Wells Fargo

AI Agents · LLM Applications

Core Components of LLM Agent Memory: What Each Stores and How It Is Written and Read

Companies: Snowflake

AI Agents · LLM Applications

Critique a CNN-Based Bridge Safety Proposal

Companies: Gsa

Daily Temperatures by Town: Linear Models to Predict NYC Temperature

Companies: Two Sigma

Regression & Statistical Modeling

Debug a From-Scratch Logistic Regression: Sigmoid Derivative and Gradients

Companies: LinkedIn

Debugging · ML Coding

Debug an inaccurate patient-message triage classifier service and write up fixes

Companies: Ro.Co

Debugging · Model Evaluation

Debug Binary Logistic Regression Training Code

Companies: LinkedIn