eth-siplab/Shift-Invariant_Deep_Learning_on_Time_Series

The official implementation of ICLR2025 paper for shift-invariant neural networks

15
/ 100
Experimental

This project helps data scientists, machine learning engineers, and researchers working with time series data improve the reliability and accuracy of their deep learning models. It takes raw time series data, processes it through a novel transformation, and outputs predictions that are consistent even if the timing of events in the input data shifts. This means your models will perform better on tasks like activity recognition, heart rate prediction, and disease classification.

No commits in the last 6 months.

Use this if your deep learning models for time series data are sensitive to slight shifts in the timing of events and you need to achieve consistent performance regardless of these shifts.

Not ideal if your primary focus is on image processing or other data types, as this solution is specifically designed for time series analysis.

time-series-analysis activity-recognition health-monitoring predictive-modeling deep-learning-research
No License Stale 6m No Package No Dependents
Maintenance 2 / 25
Adoption 5 / 25
Maturity 8 / 25
Community 0 / 25

How are scores calculated?

Stars

11

Forks

Language

Python

License

Last pushed

Apr 27, 2025

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/eth-siplab/Shift-Invariant_Deep_Learning_on_Time_Series"

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