Hacker News new | past | comments | ask | show | jobs | submit login
Improving seasonal forecast using probabilistic deep learning (arxiv.org)
59 points by lambdamore 12 days ago | hide | past | favorite | 2 comments

Specifically, using a variational autoencoder.

It's great to see that models from the deep learning literature are being adopted in fields such as weather forecasting!

Researchers determined that a NN of saliency maps could predict El NiƱo/Southern Oscillation (ENSO) in climate simulations based on relatively few observational records (page 2).

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact