AaltoML/PeriodicBNN

Code for 'Periodic Activation Functions Induce Stationarity' (NeurIPS 2021)

30
/ 100
Emerging

This project offers a specialized approach for building Bayesian neural networks that can make more reliable predictions when dealing with data that has repeating patterns or is consistent over time. It allows researchers and machine learning practitioners to input their data and specify different 'periodic activation functions' to generate models with improved stability and predictable behavior. This helps in fields where understanding data uncertainty and stationarity is crucial.

No commits in the last 6 months.

Use this if you are a machine learning researcher or practitioner working with Bayesian neural networks and need to ensure your models handle stationary or periodic data patterns effectively for more reliable predictions and uncertainty estimates.

Not ideal if you are looking for a general-purpose, off-the-shelf machine learning tool without a focus on the theoretical underpinnings of Bayesian neural networks or the specific benefits of periodic activation functions.

Bayesian-modeling neural-networks time-series-analysis uncertainty-quantification regression-classification
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 6 / 25
Maturity 16 / 25
Community 8 / 25

How are scores calculated?

Stars

19

Forks

2

Language

Jupyter Notebook

License

MIT

Last pushed

Oct 27, 2021

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/AaltoML/PeriodicBNN"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.