JonRebane / timeseries-explainability Public

forked from isaksamsten/tsexplain

Notifications You must be signed in to change notification settings
Fork 0
Star 0

Locally and globally explainable time series tweaking: New algorithms to make black‑box ML classifiers interpretable by showing the minimal changes needed to flip a prediction. These algorithms provide actionable explanations, helping practitioners understand both individual decisions and overall model behaviour.

link.springer.com/article/10.1007/s10115-019-01389-4

0 stars 4 forks Branches Tags Activity

Notifications

Name		Name	Last commit message	Last commit date
Latest commit History 8 Commits
TSC Problems		TSC Problems
tstransform		tstransform
README.md		README.md
TimeSeries_Transform.ipynb		TimeSeries_Transform.ipynb
experiment_1.py		experiment_1.py
experiment_2.py		experiment_2.py
explore_data.py		explore_data.py

Repository files navigation

This repository is outdated

For the latest implementation see Wildboar

About

Locally and globally explainable time series tweaking: New algorithms to make black‑box ML classifiers interpretable by showing the minimal changes needed to flip a prediction. These algorithms provide actionable explanations, helping practitioners understand both individual decisions and overall model behaviour.

link.springer.com/article/10.1007/s10115-019-01389-4

machine-learning time-series time-series-analysis explainable-ai explainable-ml

Report repository

Releases

No releases published

Packages

No packages published

Languages

Jupyter Notebook 80.2%
Python 17.9%
TeX 1.9%