FERMAT: Feature Engineering with Grammatical Evolution

Created by W.Langdon from gp-bibliography.bib Revision:1.8120

@InProceedings{conf/epia/Monteiro0P21,

• author = "Mariana Monteiro and Nuno Lourenco and Francisco B. Pereira",
• title = "{FERMAT}: Feature Engineering with Grammatical Evolution",
• booktitle = "Progress in Artificial Intelligence - 20th EPIA Conference on Artificial Intelligence",
• year = "2021",
• editor = "Goreti Marreiros and Francisco S. Melo and Nuno Lau and Henrique Lopes Cardoso and Luis Paulo Reis",
• volume = "12981",
• series = "Lecture Notes in Computer Science",
• pages = "239--251",
• address = "Virtual Event",
• month = sep # " 7-9",
• publisher = "Springer",
• keywords = "genetic algorithms, genetic programming, grammatical evolution, feature engineering, drug development",
• isbn13 = "978-3-030-86229-9",
• bibdate = "2021-09-20",
• bibsource = "DBLP, http://dblp.uni-trier.de/db/conf/epia/epia2021.html#Monteiro0P21",
• DOI = "doi:10.1007/978-3-030-86230-5_19",
• abstract = "Feature engineering is a key step in a machine learning study. We propose FERMAT, a grammatical evolution framework for the automatic discovery of an optimal set of engineered features, with enhanced ability to characterise data. The framework contains a grammar specifying the original features and possible operations that can be applied to data. The optimisation process searches for a transformation strategy to apply to the original dataset, aiming at creating a novel characterisation composed by a combination of original and engineered attributes. FERMAT was applied to two real-world drug development datasets and results reveal that the framework is able to craft novel representations for data that foster the predictive ability of tree-based regression models.",

}

Genetic Programming entries for Mariana Monteiro Nuno Lourenco Francisco Jose Baptista Pereira

Citations