MIFuGP: Boolean network inference from multivariate time series using fuzzy genetic programming

Created by W.Langdon from gp-bibliography.bib Revision:1.8414

@Article{Liu:2024:ins,

• author = "Xiang Liu and Yan Wang and Shan Liu and Zhicheng Ji and Shan He",
• title = "{MIFuGP:} Boolean network inference from multivariate time series using fuzzy genetic programming",
• journal = "Information Sciences",
• year = "2024",
• volume = "680",
• pages = "121129",
• keywords = "genetic algorithms, genetic programming, Boolean network inference, Multivariate time series, Fuzzy logic control, Mutual information",
• ISSN = "0020-0255",
• URL = "https://www.sciencedirect.com/science/article/pii/S0020025524010430",
• DOI = "doi:10.1016/j.ins.2024.121129",
• abstract = "Boolean network inference is essential for gaining insights into gene regulatory networks through multivariate gene expression time series. However, most existing algorithms cannot accurately reconstruct large-scale Boolean networks due to the complex and diverse relationships among genes and the overfitting problem. To address these problems, a novel inference algorithm using a mutual information-based fuzzy genetic programming approach (MIFuGP) is proposed to infer large-scale Boolean networks accurately. To represent complex regulatory relationships in Boolean networks, MIFuGP encodes Boolean functions as syntax tree programs. Taking the dependency between genes into account, MIFuGP fully extracts the mutual information from the syntax trees to alleviate the bloat problem. MIFuGP also provides a novel fitness function to make full use of state-transitions and topology information, together with a fuzzy logic control strategy to reduce the overfitting problem. Extensive experiments validate that MIFuGP significantly outperforms state-of-the-art algorithms on both real-world gene regulatory networks and artificial Boolean networks",

}

Genetic Programming entries for Xiang Liu Yan Wang Shan Liu Zhicheng Ji Shan He

Citations