Quantile Regression Workflows
summary
Quantile regression workflows use the Python package regression iser to specify pipelines that fit multiple quantiles, count points under each curve, and produce residuals for outlier detection. Examples include weather and financial time series, reconstruction of conditional CDFs, and forward simulations. Live code fits and diagnostics run quickly and show heteroscedastic behavior; an interactive Shiny-style GUI demonstrates controls, though the Python port struggles near the end of the demo.
AI-written summary · about
chapters (17)
- 0:00Quantile regression workflows
- 1:34What is quantile regression
- 3:21History and framework
- 5:58Basic example
- 7:44B-spline basis regression
- 9:32Weather data pipeline
- 11:58Heteroscedastic data handling
- 13:02Outlier detection
- 15:27Contextual outlier identification
- 16:54Conditional CDF computation
- 18:41Time series simulation
- 19:24Interactive demo setup
- 20:50Model evaluation
- 21:54Variance anomaly detection
- 23:00Point outlier methods
- 24:07NLP to regression pipelines
- 26:20Conclusion and Q&A
related talks
34mPrediction intervals
27mModelling the extreme using Quantile Regression
19mUnderstanding, Generating, and Evaluating Prediction Intervals
33mKickstart Your Probabilistic Forecasting
41mReliably forecasting time-series in real-time
01h 26mNew Trends in Estimation and Inference
30mConformal Prediction: uncertainty quantification to humanise models
33mAdaptive Prediction Intervals