Using PSO and RST to Predict the Resistant Capacity of Connections in Composite Structures

Yaima Filiberto ^[1] ; Rafael Bello ^[2] ; Yaile Caballero ^[1] ; Rafael Larrua ^[1]
1. [1] Universidad de Camagüey
  
  Universidad de Camagüey
  
  Cuba
2. [2] Universidad Central de Villa Clara
Localización: Nature Inspired Cooperative Strategies for Optimization (NICSO 2010) / coord. por Juan R. González, David Alejandro Pelta Mochcovsky , Carlos Cruz, Germán Terrazas, Natalio Krasnogor , 2010, ISBN 978-3-642-12537-9, págs. 359-370
Idioma: inglés
Enlaces
- Texto Completo Libro

Resumen
- In this paper, a method is proposed that combines the methaheuristic Particle Swarm Optimization (PSO) with the Rough Set Theory (RST) to carry out the prediction of the resistant capacity of connectors (Q) in the branch of Civil Engineering. The k-NN method is used to calculate this value. A feature selection process is performed in order to develop a more efficient process to recover the similar cases; in this case, the feature selection is done by finding the weights to be associated with the predictive features that appear in the weighted similarity function used for recovering. In this paper we propose a new alternative for calculating the weights of the features based on extended RST to the case of continuous decision features.Experimental results show that the algorithm k-NN, PSO and the method for calculating the weight of the attributes constitute an effective technique for the function approximation problem.