Multi-strategy learning made easy

• Tipo: Artigo em Livro de Atas de Conferência Internacional
• Data: 2006

Autores

Francisco Reinaldo (Autor) FEUP; Marcus Siqueira (Autor) Outra; Rui Camacho (Autor) FEUP; Luís Paulo Reis (Autor) FEUP

Ata de Conferência Internacional

Título: 10th WSEAS International Conference on Computers

Páginas: 278-284

10th WSEAS International Conference on Computers

Athens, Greece, July 13-15

Classificação Científica

FOS: Ciências da engenharia e tecnologias > Engenharia electrotécnica, electrónica e informática

CORDIS: Ciências Tecnológicas > Engenharia > Engenharia do conhecimento

Outras Informações

Abstract (EN): This paper presents the AFRANCI tool for the development of Multi-Strategy learning systems. Designing a Multi-Strategy system using AFRANCI is a two step process. The user interactively designs the structure of the system and then chooses the learning strategies for each module. After providing the datasets all modules as automatically trained. The system is aware and takes into consideration the inter-dependency of the modules. The tool has built-in learning algorithms but can use external programs implementing the learning algorithms. The tool has the following facilities. It allows any user to design in an interactive and easy fashion the structure of the target system. The structure of the target system is a collection of interconnected modules. The user may then choose the different learning algorithms to construct each module. The tool has several built-in Machine Learning algorithms has interfaces that enables it to use external learning tools like WEKA and CN2. AFRANCI uses the interdependency of the modules to determine the sequence of training. For each module the system uses a wrapper to tune automatically the parameters of the learning algorithm. In the final step of the design sequence AFRANCI generates a compact and legible ready-to-use ANSI C open-source code for the final system. To illustrate the concept we have empirically evaluated the tool in the context of the RoboCup Rescue domain. We have developed a small system that uses both neural networks, decision trees and rule induction in the same system. The experiment have shown that a very significant speed up is attained in the development of systems when using this tool.

Idioma: Inglês

Tipo (Avaliação Docente): Científica

Nº de páginas: 7