LEMON: Alternative Sampling for More Faithful Explanation Through Local Surrogate Models

Research output: Contribution to conferencePaperAcademic

3 Citations (Scopus)

Abstract

Local surrogate learning is a popular and successful method for machine learning explanation. It uses synthetic transfer data to approximate a complex reference model. The sampling technique used for this transfer data has a significant impact on the provided explanation, but remains relatively unexplored in literature. In this work, we explore alternative sampling techniques in pursuit of more faithful and robust explanations, and present LEMON: a sampling technique that samples directly from the desired distribution instead of reweighting samples as done in other explanation techniques (e.g., LIME). Next, we evaluate our technique in a synthetic and UCI dataset-based experiment, and show that our sampling technique yields more faithful explanations compared to current state-of-the-art explainers.
Original languageEnglish
Pages77-90
Number of pages14
DOIs
Publication statusPublished - 1 Apr 2023
Event21st International Symposium on Intelligent Data Analysis - Louvain-la-Neuve, Belgium
Duration: 12 Apr 202314 Apr 2023
https://ida2023.org

Conference

Conference21st International Symposium on Intelligent Data Analysis
Abbreviated titleIDA 2023
Country/TerritoryBelgium
CityLouvain-la-Neuve
Period12/04/2314/04/23
Internet address

Funding

FundersFunder number
Nederlandse Organisatie voor Wetenschappelijk Onderzoek613.009.122, 628.003.001

    Keywords

    • Machine learning
    • Interpretability
    • XAI
    • LIME
    • Sampling
    • Explainable AI

    Fingerprint

    Dive into the research topics of 'LEMON: Alternative Sampling for More Faithful Explanation Through Local Surrogate Models'. Together they form a unique fingerprint.

    Cite this