Journal
INFORMATION SCIENCES
Volume 609, Issue -, Pages 766-780Publisher
ELSEVIER SCIENCE INC
DOI: 10.1016/j.ins.2022.07.061
Keywords
Ensemble methods; Auto -ML; Grammatical evolution; Supervised learning
Categories
Funding
- University of Alicante
- Ministry of Science and Innovation of the Spanish Government
- Fondo Europeo de Desarrollo Regional (FEDER)
- Generalitat Valenciana (Conselleria d'Educacio, Investigacio, Cultura i Esport) [CIPROM/2021/21, PID2021-122263OB-C22, RTI2018-094653-B-C21/C22, PID2021-123956OB-I00]
- University of Havana
Ask authors/readers for more resources
This paper presents a two-phase optimization system that utilizes Auto-ML tools to solve classification problems and generate more robust classifiers. The experimental results show that ensembling a subset of already tested models can build a better solution, and ensuring diversity using the double-fault measure produces better results.
Automatic Machine Learning (Auto-ML) tools enable the automatic solution of real-world problems through machine learning techniques. These tools tend to be more time consum-ing than standard machine learning libraries, therefore, exploiting all the available resources to the full is a valuable feature. This paper presents a two-phase optimization system for solving classification problems. The system is designed to produce more robust classifiers by exploiting the different architectures that are generated while solving classi-fication problems with Auto-ML tools, particularly AutoGOAL. In the first phase, the system follows a probabilistic strategy to find the best combination of algorithms and hyperpa-rameters to generate a collection of base models according to certain diversity criteria; and in the second, it follows similar Auto-ML strategies to ensemble those models. The HAHA 2019 challenge corpus and the Adult dataset were used to evaluate the system. The experimental results show that: i) a better solution can be built by ensembling a subset of the already tested models; ii) the performance of ensemble methods depends on the col-lection of base models used; and, iii) ensuring diversity using the double-fault measure produces better results than the disagreement measure. The source code is available online for the research community. (c) 2022 Elsevier Inc. All rights reserved.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available