Prototype-based Explainable Neural Networks with Channel-specific Reasoning for Geospatial Learning Tasks

TL;DR

This paper introduces a prototype-based explainable neural network tailored for multi-channel geospatial data, enabling interpretable predictions by identifying channel-specific prototypes, demonstrated through climate and satellite imagery classification tasks.

Contribution

The study develops a novel prototype-based XAI method optimized for multi-channel geoscientific data, allowing channel-specific reasoning and improving interpretability without sacrificing performance.

Findings

Achieved comparable accuracy to standard neural networks on geoscientific tasks.

Provided both local and global explanations highlighting feature relevance across channels.

Enhanced model transparency and trustworthiness in geospatial learning applications.

Abstract

Explainable AI (XAI) is essential for understanding machine learning (ML) decision-making and ensuring model trustworthiness in scientific applications. Prototype-based XAI methods offer an intrinsically interpretable alternative to post-hoc approaches which often yield inconsistent explanations. Prototype-based XAI methods make predictions based on the similarity between inputs and learned prototypes that represent typical characteristics of target classes. However, existing prototype-based models are primarily designed for standard RGB image data and are not optimized for the distinct, variable-specific channels commonly found in geoscientific image and raster datasets. In this study, we develop a prototype-based XAI approach tailored for multi-channel geospatial data, where each channel represents a distinct physical environmental variable or spectral channel. Our approach enables the model to identify separate, channel-specific prototypical characteristics sourced from multiple distinct training examples that inform how these features individually and in combination influence model prediction while achieving comparable performance to standard neural networks. We demonstrate this method through two geoscientific case studies: (1) classification of Madden Julian Oscillation phases using multi-variable climate data and (2) land-use classification from multispectral satellite imagery. This approach produces both local (instance-level) and global (model-level) explanations for providing insights into feature-relevance across channels. By explicitly incorporating channel-prototypes into the prediction process, we discuss how this approach enhances the transparency and trustworthiness of ML models for geoscientific learning tasks.