antoinecarme / pyaf

PyAF is an Open Source Python library for Automatic Time Series Forecasting built on top of popular pydata modules.
BSD 3-Clause "New" or "Revised" License
458 stars 73 forks source link

Investigate probabilistic forecasting #140

Closed antoinecarme closed 3 years ago

antoinecarme commented 4 years ago

It would be nice to have some kind of probabilistic forecasting in PyAF (version x.0)

For the moment, PyAF is predicting the future values of the signal (point forecasts). Probabilistic forecasting aims at producing forecasts for the whole distribution over the horizon instead of just producing point estimates.

Roughly speaking, Probabilistic forecasting is about producing estimates of all forecast quantiles over the horizon.

antoinecarme commented 4 years ago

Probabilistic Forecasting with Spline Quantile Function RNNs

Jan Gasthaus Konstantinos Benidis Yuyang Wang Syama S. RangapuramDavid Salinas Valentin Flunkert Tim JanuschowskiAWS AI Labs

http://proceedings.mlr.press/v89/gasthaus19a/gasthaus19a.pdf

antoinecarme commented 4 years ago

Linked with this issue :

Evaluate Continuous Ranked Probability Score as a Forecasting Performance Measure #74

antoinecarme commented 4 years ago

ensembleBMA: An R Package for Probabilistic Forecasting using Ensembles and Bayesian Model Averaging

Chris Fraley, Adrian E. RafteryTilmann Gneiting, J. McLean Sloughter

https://www.stat.washington.edu/sites/default/files/files/reports/2008/tr516.pdf

antoinecarme commented 4 years ago

Nice doc with sample cases/datasets

https://ubc-mds.github.io/DSCI_562_regr-2/lec4/032-probabilistic_forecasting.html

antoinecarme commented 4 years ago

A note on the use of empirical AUC for evaluating probabilistic forecasts

Simon Byrne

https://projecteuclid.org/download/pdfview_1/euclid.ejs/1455715967

antoinecarme commented 3 years ago

Closing