Explaining Deep Classification of Time-Series Data with Learned Prototypes

TL;DR

This paper introduces a prototype-based explainability method for deep learning models classifying 2D time-series data, enhancing interpretability by visualizing learned prototypes that correspond to real-world features across various biomedical and audio applications.

Contribution

It presents a novel prototype learning framework for 2D time-series data, improving interpretability and robustness in deep classification models across multiple domains.

Findings

Prototypes effectively learn real-world features like bradycardia, apnea, and speech articulation.

Enhanced prototype diversity improves model robustness.

Visualizations elucidate how prototypes influence classification decisions.

Abstract

The emergence of deep learning networks raises a need for explainable AI so that users and domain experts can be confident applying them to high-risk decisions. In this paper, we leverage data from the latent space induced by deep learning models to learn stereotypical representations or "prototypes" during training to elucidate the algorithmic decision-making process. We study how leveraging prototypes effect classification decisions of two dimensional time-series data in a few different settings: (1) electrocardiogram (ECG) waveforms to detect clinical bradycardia, a slowing of heart rate, in preterm infants, (2) respiration waveforms to detect apnea of prematurity, and (3) audio waveforms to classify spoken digits. We improve upon existing models by optimizing for increased prototype diversity and robustness, visualize how these prototypes in the latent space are used by the model to distinguish classes, and show that prototypes are capable of learning features on two dimensional time-series data to produce explainable insights during classification tasks. We show that the prototypes are capable of learning real-world features - bradycardia in ECG, apnea in respiration, and articulation in speech - as well as features within sub-classes. Our novel work leverages learned prototypical framework on two dimensional time-series data to produce explainable insights during classification tasks.