Semifinal results of a research project involving algorithmic complexity estimation and machine learning

Abstract

In recent years, the intersection of algorithmic complexity and machine learning has opened new avenues for analyzing continuous biomedical data. This paper presents the semifinal results of a research project focused on estimating Kolmogorov-Chaitin Complexity (KCC) using the Block Decomposition Method. KCC serves as a core feature for machine learning models aimed at predicting sepsis and epileptic seizures. The results highlight the efficacy of these models, with promising performance metrics, and underscore the utility of algorithmic complexity measures in enhancing machine learning models for biomedical applications. By leveraging the inherent complexity in biomedical signals, these models achieve superior predictive performance.