Explainable Matrix - Visualization for Global and Local Interpretability of Random Forest Classification Ensembles



Mário Popolin Neto, Fernando Paulovich

 External link (DOI) 

 View presentation:2020-10-28T14:30:00ZGMT-0600Change your timezone on the schedule page
2020-10-28T14:30:00Z

Exemplar figure — Explainable Matrix (ExMatrix) is a method for Random Forest (RF) interpretability based on logic rules' visual representations. ExMatrix supports global and local explanations enabling the analysis of RF models and the auditing of classification results. The key idea is to explore logic rules using a matrix-like metaphor, where rows are rules, columns are features, and cells are rules predicates. ExMatrix allows reasoning on a considerable number of rules at once, helping users to build insights by employing different orderings of rules/rows and features/columns, supporting the overview of entire RF models while also enables focusing on specific parts for details on-demand.

Fast forward

Direct link to video on YouTube: https://youtu.be/qlthySP_mwA

Keywords

Random forest visualization, logic rules visualization, classification model interpretability, explainable artificial intelligence

Abstract

Over the past decades, classification models have proven to be essential machine learning tools given their potential and applicability in various domains. In these years, the north of the majority of the researchers had been to improve quality metrics, notwithstanding the lack of information about models' decisions such metrics convey. Recently, this paradigm has shifted, and strategies beyond tables and numbers to assist in interpreting models' decisions are increasing in importance. Part of this trend, visualization techniques have been extensively used to support the interpretability of classification models, with a significant focus on rule-based techniques. Despite the advances, the existing approaches present limitations in terms of visual scalability, and the visualization of large and complex models, such as the ones produced by the Random Forest (RF) technique, remains a challenge. In this paper, we propose Explainable Matrix (ExMatrix), a novel visualization method for RF interpretability that can handle models with massive quantities of rules. It employs a simple yet powerful matrix-like visual metaphor, where rows are rules, columns are features, and cells are rules predicates, enabling the analysis of entire models and auditing classification results. ExMatrix applicability is confirmed via different examples, showing how it can be used in practice to promote RF models interpretability.