XAI Method for explaining time-series

LASTS is a framework that can explain the decisions of black box models for time series classification. The explanation consists of factual and counterfactual rules revealing the reasons for the classification through conditions expressed as subsequences that must or must not be contained in the time series. In addition, a set of exemplar and counterexemplar time series highlight similarities and differences with the time series under analysis. A wide experimentation shows that the proposed method provides faithful, meaningful, stable, and interpretable explanations. Besides, we show that the proposed explainer can be employed on multivariate time series.

Data and Resources
To access the resources you must log in

This item has no data

Additional Info
Field Value
Accessibility Both
AccessibilityMode OnLine Access
Availability On-Line
Basic rights Download
CreationDate 2021-05-17
Creator Spinnato, Francesco,
Field/Scope of use Non-commercial research only
Group Social Impact of AI and explainable ML
Owner Spinnato, Francesco,
ProgrammingLanguage Python
Sublicense rights No
Territory of use World Wide
Thematic Cluster Other
system:type Method
Management Info
Field Value
Author Spinnato Francesco
Maintainer Spinnato Francesco
Version 1
Last Updated 8 September 2023, 14:49 (CEST)
Created 17 May 2021, 17:52 (CEST)