amazon-science/unconditional-time-series-diffusion

Official PyTorch implementation of TSDiff models presented in the NeurIPS 2023 paper "Predict, Refine, Synthesize: Self-Guiding Diffusion Models for Probabilistic Time Series Forecasting"

44
/ 100
Emerging

This tool helps experts who work with time-based data to predict future trends, improve existing forecasts, or generate realistic synthetic data. You input your historical time series data, and it outputs probabilistic forecasts, refined predictions, or new synthetic datasets. It's ideal for data scientists, analysts, or researchers dealing with complex time series like electricity consumption, sales figures, or sensor readings.

248 stars. No commits in the last 6 months.

Use this if you need advanced, probabilistic forecasts for time series data, want to enhance the accuracy of your current prediction models, or require realistic synthetic data for training new models.

Not ideal if you only need simple, deterministic point forecasts or are working with non-sequential, static datasets.

time-series-forecasting predictive-analytics data-synthesis financial-modeling resource-planning
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 18 / 25

How are scores calculated?

Stars

248

Forks

36

Language

Python

License

Apache-2.0

Last pushed

Jan 30, 2024

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/amazon-science/unconditional-time-series-diffusion"

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