Lyft Demand Modeling & Forecasting Questions

Techniques for modeling and forecasting ride-hailing demand, including time-series forecasting, demand drivers, feature engineering, model selection (e.g., ARIMA, Prophet, ML-based predictors), evaluation metrics (MAPE, RMSE), and deployment considerations within analytics workflows for transportation data.

EasyTechnical

56 practiced

At a high level, contrast ARIMA and Prophet for demand forecasting on Lyft hourly data. Discuss strengths and weaknesses of each in handling seasonality, holidays/events, missing data, and scaling to thousands of zones.

MediumTechnical

61 practiced

Implement a Python function that, given arrays y_true and y_pred, computes MAPE safely (handling zeros) and symmetric MAPE (sMAPE). Explain the edge cases and why plain MAPE can be misleading for Lyft hourly demand with many near-zero values.

EasyTechnical

66 practiced

What are lag features in the context of time-series forecasting? Describe three types of lag or window-based features you would compute for predicting 15-minute demand and explain why each is useful for Lyft demand modeling.

EasyTechnical

59 practiced

Explain what probabilistic forecasting is and why it matters for Lyft demand modeling. Give two examples of decision-making tasks that require prediction intervals or quantiles instead of point forecasts.

MediumTechnical

55 practiced

Design an A/B experiment to compare two demand forecasting models in production. Define: randomization unit (zones/time), primary and secondary metrics (business KPIs and forecasting metrics), experiment duration, and safeguards to avoid harm to drivers/customers during the test.

Unlock Full Question Bank

Get access to hundreds of Lyft Demand Modeling & Forecasting interview questions and detailed answers.

Join thousands of developers preparing for their dream job.