Interactive dense pixel visualizations for time series and model attribution explanations

TL;DR

This paper introduces DAVOTS, an interactive visualization tool that enables detailed exploration of time series data, neural network activations, and attributions to improve understanding and evaluation of model explanations.

Contribution

The paper presents DAVOTS, a novel dense pixel visualization method tailored for time series explanations, incorporating clustering and ordering strategies for enhanced data exploration.

Findings

Effective visualization of CNN activations on time series data.

Clustering enhances pattern discovery in large datasets.

Facilitates better understanding of model decisions and explanations.

Abstract

The field of Explainable Artificial Intelligence (XAI) for Deep Neural Network models has developed significantly, offering numerous techniques to extract explanations from models. However, evaluating explanations is often not trivial, and differences in applied metrics can be subtle, especially with non-intelligible data. Thus, there is a need for visualizations tailored to explore explanations for domains with such data, e.g., time series. We propose DAVOTS, an interactive visual analytics approach to explore raw time series data, activations of neural networks, and attributions in a dense-pixel visualization to gain insights into the data, models' decisions, and explanations. To further support users in exploring large datasets, we apply clustering approaches to the visualized data domains to highlight groups and present ordering strategies for individual and combined data exploration to facilitate finding patterns. We visualize a CNN trained on the FordA dataset to demonstrate the approach.