Approximate LTL model checking

TL;DR

This paper introduces a machine learning-based approximate method for LTL model checking that significantly reduces computational time while maintaining high accuracy, addressing the state explosion problem.

Contribution

It proposes a novel ML approach using Boosted Tree to predict LTL model checking results, offering a scalable alternative to traditional methods.

Findings

Achieves 98% average accuracy in predictions.

Provides up to 9.4 million times efficiency improvement.

Effectively handles LTL formulas of length 500.

Abstract

The state explosion problem and the exponentially computational complexity restrict the further applications of LTL model checking. To this end, this study tries to seek an acceptable approximate solution for LTL model checking by introducing the Machine Learning (ML) technique, and a method for predicting results of LTL model checking via the Boosted Tree (BT) algorithm is proposed in this paper. First, for a number of Kripke structures and LTL formulas, a data set A containing their model checking results is obtained, using the existing LTL model checking algorithm. Second, the LTL model checking problem can be induced to a binary classification problem of machine learning. In other words, some records in A form a training set for the BT algorithm. On the basis of it, a ML model M is obtained to predict the results of LTL model checking. As a result, an approximate LTL model checking technique occurs. The experiments show that the new method has the average accuracy of 98.0%, and its average efficiency is 9.4 million times higher than that of the representative model checking method, if the length of each of LTL formulas equals to 500.