A predictive system resilience assessment tool (PSRAT)

This paper presents the Predictive System Resilience Assessment Tool (PSRAT) developed to automatically apply regression methods to accurately predict system resilience in terms of the external activities that impact system performance. The tool enables visualization of model fit with different covariates to the performance data, and computation of goodness-of-fit, resilience metrics, and confidence intervals to guide decision-making. The application and source code are freely available from the GitHub repository dedicated to this project. The abstract mathematical nature of the models enables resilience assessment and prediction in a range of application domains.