Machine Learning interview questions

Showing 511–540 of 572 results

• Perform no-intercept linear regression from two datasetsMedium

  This question evaluates linear regression modeling, numerical linear algebra (normal equations and stable factorization methods), regularization, R² interpretation, and data-preprocessing/validation skills within the Machine Learning domain for a Data Scientist role.

  Two SigmaData ScientistMachine LearningOnline Assessment
• Predict Stock Prices from Google Search DataMedium

  This question evaluates a candidate's competency in time-series and cross-sectional predictive modeling, alternative-data feature engineering, and rigorous validation to prevent temporal leakage within the Machine Learning domain.

  Two SigmaData ScientistMachine LearningTechnical Screen
• Predict a Company's Stock Price from Six Years of Web Search LogsHard

  This question evaluates proficiency in feature engineering and time-series forecasting from large-scale event logs, including entity linking, aggregation and normalization, handling non-stationarity and drift, and model selection and validation for predicting a given company's stock price.

  Two SigmaData ScientistMachine LearningTechnical Screen
• Predict bike demand and avoid overfittingHard

  This question evaluates time-series forecasting, feature engineering, awareness of data leakage risks, model evaluation choices, and overfitting prevention competencies within the Machine Learning domain applied to demand prediction.

  Two SigmaData ScientistMachine LearningTechnical Screen
• Slopes of y-on-x vs x-on-y Regression: Relationship and When They Are EqualHard

  This question evaluates understanding of simple OLS linear regression estimators, moment-based expressions for slopes, the relationship between y-on-x and x-on-y fitted slopes, and how correlation and variable scaling influence coefficient values.

  Two SigmaData ScientistMachine LearningTechnical Screen
• Why Is the Sample Mean Approximately Normal for Large Samples?Medium

  This question evaluates understanding of the Central Limit Theorem and asymptotic statistical reasoning, including recognition that the standardized sample mean (equivalently the standardized sum) becomes approximately normal and how that fact underpins inference such as confidence intervals.

  Two SigmaData ScientistMachine LearningTechnical Screen
• Build cold-start restaurant ratingsMedium

  This problem tests the ability to design an end-to-end machine learning system for predicting restaurant ratings when historical user-item interactions are scarce, focusing on feature engineering, model selection, and bias mitigation.

  UberData ScientistMachine LearningTechnical Screen
• Design ETA prediction for Uber ridesHard

  This question evaluates competency in end-to-end machine learning system design for real-time spatiotemporal ETA prediction, encompassing feature engineering from heterogeneous data sources, uncertainty quantification, model selection, and production serving and monitoring.

  UberData ScientistMachine LearningTechnical Screen
• Design a Ride-Hailing ETA SystemMedium

  This question evaluates a data scientist's competency in end-to-end machine learning system design for spatiotemporal prediction, covering temporal and geospatial feature engineering, model family selection, metric tradeoffs and calibration, selection bias and noise handling, latency and deployment constraints, monitoring, and experiment design.

  UberData ScientistMachine LearningTechnical Screen
• Evaluate Promotions for Uber Eats UsersMedium

  This question tests the ability to design an A/B test for promotions in a two-sided marketplace, focusing on identifying interference (network effects) and selecting appropriate metrics to measure causal impact.

  UberData ScientistMachine LearningTechnical Screen
• Explain and test completion-rate gapsEasy

  This question evaluates understanding of causal inference, product-metrics analysis, experimental design, and statistical hypothesis testing (including two-sample t-tests) as applied to operational problems in a marketplace.

  UberMachine Learning EngineerMachine LearningTechnical Screen
• How would you build UberEats ranking?Medium

  This question assesses the ability to design an end-to-end ranking system for a food delivery platform, focusing on objective formulation, feature selection, and the trade-offs between relevance and business constraints.

  UberData ScientistMachine LearningTechnical Screen
• How would you design an ETA prediction system?Easy

  This problem tests the ability to design an end-to-end machine learning system for predicting ride arrival times, focusing on feature selection, model architecture, and evaluation strategies.

  UberData ScientistMachine LearningTechnical Screen
• Implement 1D convex minimization in PythonMedium

  An Uber Machine Learning Engineer technical-screen question on derivative-free 1D convex optimization: minimize a black-box convex function on a closed interval using only function evaluations. It tests method selection (golden-section vs ternary search), stopping criteria and tolerances, function-evaluation and time/space complexity analysis, and robust handling of flat regions, boundary optima, noisy evaluations, and numerical precision.

  UberMachine Learning EngineerMachine LearningTechnical Screen
• Implement CLIP Contrastive LossMedium

  This question evaluates understanding and implementation of contrastive representation learning concepts—specifically similarity matrices, symmetric image-text contrastive loss, normalization, temperature scaling, and label construction—testing competency in building losses for embedding alignment in the Machine Learning domain.

  UberMachine Learning EngineerMachine LearningTechnical Screen
• Implement Multi-Head Self-AttentionMedium

  This question tests the ability to implement the multi-head self-attention mechanism from scratch using PyTorch, focusing on tensor manipulation, linear projections, and the scaled dot-product attention formula.

  UberMachine Learning EngineerMachine LearningTechnical Screen
• Implement Streaming Clustering for NumbersMedium

  This question evaluates a data scientist's competency in online clustering for numeric streams, testing understanding of streaming algorithms, bounded-memory cluster summaries, online updates, and issues including initialization, convergence behavior, outlier handling, and concept drift.

  UberData ScientistMachine LearningOnsite
• Implement linear and logistic regressionMedium

  This question evaluates understanding and hands-on implementation of linear and logistic regression, testing competencies in prediction functions, loss formulation, gradient computation, gradient descent training, regularization, and model selection within the Machine Learning domain (supervised learning and statistical modeling).

  UberMachine Learning EngineerMachine LearningTechnical Screen
• Explain Core ML FundamentalsEasy

  This question tests your conceptual understanding of how fundamental components of neural networks and predictive models affect optimization, generalization, and performance.

  UiPathMachine Learning EngineerMachine LearningTechnical Screen
• Explain Transformer Layers and FFN RationaleMedium

  This interview question evaluates core ML concepts, assumptions, math intuition, training/evaluation trade-offs, and practical failure modes in a realistic interview setting. A strong answer for Explain Transformer Layers and FFN Rationale states assumptions, handles edge cases, explains trade-offs, and shows how to validate the result clearly.

  UiPathMachine Learning EngineerMachine LearningTechnical Screen
• LLM fundamentals: Transformer internals, beam search decoding, and LoRA fine-tuningMedium

  An LLM fundamentals question that asks you to explain Transformer internals such as scaled self-attention, multi-head attention, positional information and causal masking. It also covers how beam search decoding compares with greedy decoding and sampling, and how LoRA reduces the cost of fine-tuning.

  UiPathMachine Learning EngineerMachine LearningTechnical Screen
• Reduce overfitting and explain the mean squared error lossMedium

  A machine learning phone screen on how to diagnose and reduce overfitting, and on the mean squared error loss: its definition, what minimizing it estimates, its probabilistic interpretation, and when another loss works better. Tests fundamentals, mechanisms and trade-off reasoning.

  UiPathMachine Learning EngineerMachine LearningHR Screen
• Reliable LLM tool calling: format control, failure recovery, and runaway loopsMedium

  An LLM agent question on tool calling: how a model invokes tools, how to keep tool calls in a valid format, how to make tool use stable and fix wrong calls, and how to handle tool-call loops that run too long. It tests practical reliability engineering for production agents.

  UiPathMachine Learning EngineerMachine LearningTechnical Screen
• Derive logistic regression objective and gradientsEasy

  This question evaluates understanding of logistic regression, maximum likelihood estimation, cross-entropy (negative log-likelihood), L2 regularization, and the ability to compute model gradients for binary classification.

  UpstartData ScientistMachine LearningHR Screen
• Design Push-Notification System for Airport Surge PricingMedium

  Design Push-Notification System for Airport Surge Pricing evaluates core ML concepts, assumptions, math intuition, training/evaluation trade-offs, and practical failure modes in a realistic interview setting. A strong answer states assumptions, handles edge cases, explains trade-offs, and shows how to validate the result clearly.

  UpstartData ScientistMachine LearningTechnical Screen
• Design a Real-Time Personalized Ad Selection SystemHard

  This interview question evaluates core ML concepts, assumptions, math intuition, training/evaluation trade-offs, and practical failure modes in a realistic interview setting. A strong answer for Design a Real-Time Personalized Ad Selection System states assumptions, handles edge cases, explains trade-offs, and shows how to validate the result clearly.

  UpstartData ScientistMachine LearningTechnical Screen
• Explain L1 vs L2 and ridge vs lassoEasy

  This question evaluates understanding of regularization techniques in Machine Learning—specifically distinctions between L1 and L2 norms and their instantiation as Lasso and Ridge regression—covering effects on objective functions, parameter sparsity, geometric intuitions, handling of correlated features, and the role of feature scaling.

  UpstartData ScientistMachine LearningOnsite
• How to Architect a Personalized Ads Serving SystemHard

  This interview question evaluates core ML concepts, assumptions, math intuition, training/evaluation trade-offs, and practical failure modes in a realistic interview setting. A strong answer for How to Architect a Personalized Ads Serving System states assumptions, handles edge cases, explains trade-offs, and shows how to validate the result clearly.

  UpstartData ScientistMachine LearningTechnical Screen
• Leverage Existing Model for Low Credit Score ApplicantsMedium

  Leverage Existing Model for Low Credit Score Applicants evaluates core ML concepts, assumptions, math intuition, training/evaluation trade-offs, and practical failure modes in a realistic interview setting. A strong answer states assumptions, handles edge cases, explains trade-offs, and shows how to validate the result clearly.

  UpstartData ScientistMachine LearningOnsite
• Assemble a RAG Pipeline From Provided Python Interfaces and Explain RetrievalMedium

  Assemble a retrieval-augmented generation pipeline from provided Python interfaces, covering chunking, embedding, vector search and grounded prompting, then explain why embeddings and cosine similarity work for retrieval and when RAG beats fine-tuning or long context. It tests applied LLM engineering and clear reasoning with unfamiliar tooling.

  VisaSoftware EngineerMachine LearningOnsite