Describe two methods for generating prediction intervals or probabilistic forecasts
Tests uncertainty quantification for risk-adjusted decisions. Strong answers: (1) parametric intervals via forecast error variance and normal multipliers, (2) bootstrap residual resampling for empirical percentiles.
Tests ability to measure forecast uncertainty for business risk. Good answers cover two distinct valid approaches: (1) parametric intervals via point forecast plus or minus a multiplier times estimated h-step standard deviation, assuming normal errors; (2) bootstrap methods that resample residuals to generate empirical percentiles without normality assumptions. Note that variance typically grows with increasing forecast horizon. Red flag: confusing prediction intervals with confidence intervals or ignoring horizon-dependent error scaling.
Read the original → otexts.com
- #forecasting
- #prediction intervals
- #uncertainty quantification
- #time series
- #analytics
Get five bites like this every day.
Tezvyn delivers a daily feed of 60-second tech bites with quizzes to lock in what you learn.