Explainable Failure Predictions with RNN Classifiers based on Time Series Data

TL;DR

This paper presents an explainable RNN-based approach for failure prediction in storage systems using time series data, which identifies key anomalous events to provide interpretability alongside accurate predictions.

Contribution

It introduces a novel method combining event extraction and attention-based RNNs to improve explainability in failure prediction from time series telemetry data.

Findings

Predicts failures within a 3-day window with accuracy comparable to traditional methods.

Provides explanations by highlighting key anomalous events leading to failures.

Effective in real-world storage environments.

Abstract

Given key performance indicators collected with fine granularity as time series, our aim is to predict and explain failures in storage environments. Although explainable predictive modeling based on spiky telemetry data is key in many domains, current approaches cannot tackle this problem. Deep learning methods suitable for sequence modeling and learning temporal dependencies, such as RNNs, are effective, but opaque from an explainability perspective. Our approach first extracts the anomalous spikes from time series as events and then builds an RNN classifier with attention mechanisms to embed the irregularity and frequency of these events. A preliminary evaluation on real world storage environments shows that our approach can predict failures within a 3-day prediction window with comparable accuracy as traditional RNN-based classifiers. At the same time it can explain the predictions by returning the key anomalous events which led to those failure predictions.