Ponencia
The metric-aware kernel-width choice for LIME
Autor/es | Barrera Vicent, Aurelio
Paluzo Hidalgo, Eduardo Gutiérrez Naranjo, Miguel Ángel ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Departamento | Universidad de Sevilla. Departamento de Ciencias de la Computación e Inteligencia Artificial |
Fecha de publicación | 2023 |
Fecha de depósito | 2024-04-23 |
Publicado en |
|
ISBN/ISSN | 1613-0073 |
Resumen | Local Interpretable Model-Agnostic Explanations (LIME) are a well-known approach to provide local interpretability to Machine Learning models. LIME uses an exponential smoothing kernel based on the kernel width value, which ... Local Interpretable Model-Agnostic Explanations (LIME) are a well-known approach to provide local interpretability to Machine Learning models. LIME uses an exponential smoothing kernel based on the kernel width value, which defines the width of the local neighbourhood. In this paper, we study the influence of the distances for these local explanations, and we explore the choice of kernel width to guarantee a fair performance comparison between the distances. |
Cita | Barrera Vicent, A., Paluzo Hidalgo, E. y Gutiérrez Naranjo, M.Á. (2023). The metric-aware kernel-width choice for LIME. En 1st World Conference on eXplainable Artificial Intelligence: Late-Breaking Work, Demos and Doctoral Consortium, xAI-2023: LB-D-DC Lisbon 26 July 2023 through 28 July 2023 (117-122), Lisboa: CEUR-WS. |
Ficheros | Tamaño | Formato | Ver | Descripción |
---|---|---|---|---|
The metric-aware kernel-width ... | 1.013Mb | ![]() | Ver/ | |