This paper focuses on the problem of choosing one among many alternatives, each one expressed as a combination of factors. The problem is approached with a reasoning-based strategy that takes into account the relations among the alternatives discovered by means of a data mining technique. The problem of choosing a combination of different vegetables based on the synergistic effects of the combination and considering some socioeconomic variables is taken as a case study. The idea is to identify combinations that lead to gains in productivity, profitability, and lower costs. Experts recognize that some combinations of cultures can generate synergistic effects that can lead to profits or losses, depending on the production variables involved. However, the analysis of the results of the cultivation of multiple varieties involves a space of possibilities whose treatment is not trivial. Among the alternatives available, this study explored the Combinatorial Neural Model (CNM) that assures the hypotheses generation for all possible combinations, within limits defined by the model parameters. The study was carried out on data collected from farms in the Brazilian Federal District. Two approaches to the problem are presented: (i) the first one based on univariate and bi-dimensional data analysis and (ii) a multidimensional analysis based on the CNM results.
[1]
Valmir Carneiro Barbosa,et al.
Learning in the combinatorial neural model
,
1998,
IEEE Trans. Neural Networks.
[2]
R. J. Machado.
Handling knowledge in high order neural networks: the combinatorial neural model
,
1989,
International 1989 Joint Conference on Neural Networks.
[3]
Ronald R. Yager,et al.
Uncertainty in Knowledge Bases
,
1990,
Lecture Notes in Computer Science.
[4]
Ning Zhong,et al.
Methodologies for Knowledge Discovery and Data Mining
,
2002,
Lecture Notes in Computer Science.
[5]
Wolfgang Pree,et al.
Optimization of the combinatorial neural model
,
1998,
Proceedings 5th Brazilian Symposium on Neural Networks (Cat. No.98EX209).
[6]
Paulo Martins Engel,et al.
Accuracy Tuning on Combinatorial Neural Model
,
1999,
PAKDD.
[7]
Simon Haykin,et al.
Neural Networks: A Comprehensive Foundation
,
1998
.
[8]
Armando Freitas da Rocha,et al.
The Combinatorial Neural Network: A Connectionist Model for Knowledge Based Systems
,
1990,
IPMU.
[9]
Paulo Martins Engel,et al.
Alleviating the complexity of the Combinatorial Neural Model using a committee machine
,
2000
.
[10]
Hércules Antonio do Prado,et al.
A parsimonious generation of combinatorial neural model
,
1998
.