The Curious Quant cover image

EP20: Prof Rob Hyndman: Forecasting COVID, time-series, and why causality doesnt matter as much as you think.

The Curious Quant

00:00

What Are Good Practices in Forecasting?

Time series forecasting was a big part of the literature growing up around finance I suppose in the academic world. In our case it was very strongly dominated by arch-garch these type of volatility type modeling mean reversionary type of ideas. One thing is to always do some kind of evaluation on how well your forecasts have done historically and that will mean doing something like forecasting using a rolling forecast origin. Never just use one model so you always want to be trying multiple models and often then you'll want to do an ensemble because it'll be better to combine models rather than to use an individual model yeah.

Transcript
Play full episode

The AI-powered Podcast Player

Save insights by tapping your headphones, chat with episodes, discover the best highlights - and more!
App store bannerPlay store banner
Get the app