English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

This tutorial focuses on efficient methods to predictive monitoring (PM), theproblem of detecting at runtime future violations of a given requirement fromthe current state of a system. While performing model checking at runtime wouldoffer a precise solution to the PM problem, it is generally computationallyexpensive. To address this scalability issue, several lightweight approachesbased on machine learning have recently been proposed. These approaches work bylearning an approximate yet efficient surrogate (deep learning) model of theexpensive model checker. A key challenge remains to ensure reliablepredictions, especially in safety-critical applications. We review our recentwork on predictive monitoring, one of the first to propose learning-basedapproximations for CPS verification of temporal logic specifications and thefirst in this context to apply conformal prediction (CP) for rigorousuncertainty quantification. These CP-based uncertainty estimators offerstatistical guarantees regarding the generalization error of the learningmodel, and they can be used to determine unreliable predictions that should berejected. In this tutorial, we present a general and comprehensive frameworksummarizing our approach to the predictive monitoring of CPSs, examining indetail several variants determined by three main dimensions: system dynamics(deterministic, non-deterministic, stochastic), state observability, andsemantics of requirements' satisfaction (Boolean or quantitative).

Learning-Based Approaches to Predictive Monitoring with Conformal Statistical Guarantees