Flávio Joaquim de Souza

Inverted hierarchical neuro-fuzzy BSP system: a novel neuro-fuzzy model for pattern classification and rule extraction in databases

This paper introduces the Inverted Hierarchical Neuro-Fuzzy BSP System (HNFB/sup -1/), a new neuro-fuzzy model that has been specifically created for record classification and rule extraction in databases. The HNFB/sup -1/ is based on the Hierarchical Neuro-Fuzzy Binary Space Partitioning Model (HNFB), which embodies a recursive partitioning of the input space, is able to automatically generate its own structure, and allows a greater number of inputs. The new HNFB/sup -1/ allows the extraction of knowledge in the form of interpretable fuzzy rules expressed by the following: If x is A and y is B, then input pattern belongs to class Z. For the process of rule extraction in the HNFB/sup -1/ model, two fuzzy evaluation measures were defined: 1) fuzzy accuracy and 2) fuzzy coverage. The HNFB/sup -1/ has been evaluated with different benchmark databases for the classification task: Iris Dataset, Wine Data, Pima Indians Diabetes Database, Bupa Liver Disorders, and Heart Disease. When compared with several other pattern classification models and algorithms, the HNFB/sup -1/ model has shown similar or better classification performance. Nevertheless, its performance in terms of processing time is remarkable. The HNFB/sup -1/ converged in less than one minute for all the databases described in the case study.

Hierarchical neuro-fuzzy quadtree models

Hybrid neuro-fuzzy systems have been in evidence during the past few years, due to its attractive combination of the learning capacity of artificial neural networks with the interpretability of the fuzzy systems. This article proposes a new hybrid neuro-fuzzy model, named hierarchical neuro-fuzzy quadtree (HNFQ), which is based on a recursive partitioning method of the input space named quadtree. The article describes the architecture of this new model, presenting its basic cell and its learning algorithm. The HNFQ system is evaluated in three well known benchmark applications: the sinc(x, y) function approximation, the Mackey Glass chaotic series forecast and the two spirals problem. When compared to other neurofuzzy systems, the HNFQ exhibits competing results, with two major advantages it automatically creates its own structure and it is not limited to few input variables.

Electric load forecasting: evaluating the novel hierarchical neuro-fuzzy BSP model

This paper introduces a new hybrid neuro-fuzzy model, called HNFB, and evaluates its performance in short-term load forecasting. To this end, two Brazilian electric power companies were used as case studies. A total of three intelligent forecasting systems were tested and compared: neural networks, neuro-fuzzy, and neural/neuro-fuzzy systems. As input variables, the experiments made use of historical load series and of additional variables that influence the load behavior, such as the temperature, the comfort index and the consumption profile. The results reveal the potential of the proposed neuro-fuzzy and neural/neuro-fuzzy models for load forecasting, when compared with neural networks; the mean absolute percentage errors varied between 0.44 and 1.95%, depending on the case study at hand.

Reinforcement learning/spl I.bar/hierarchical neuro-fuzzy politree model for control of autonomous agents

This work presents a new hybrid neuro-fuzzy model for automatic learning of actions taken by agents. The main objective of this new model is to provide an agent with intelligence, making it capable, by interacting with its environment, to acquire and retain knowledge for reasoning (infer an action). This new model, named reinforcement learning hierarchical neuro-fuzzy politree (RL-HNFP), descends from the reinforcement learing hierarchical neuro-fuzzy BSP (RL-HNFB) that uses binary space partitioning. By using hierarchical partitioning methods, together with the reinforcement learning (RL) methodology, a new class of neuro-fuzzy systems (SNF) was obtained, which executes, in addition to automatically learning its structure, the autonomous learning of the actions to be taken by an agent. These characteristics represent an important differential when compared with the existing intelligent agents learning systems. The obtained results demonstrate the potential of this new model, which operates without any prior information, such as number of rules, rules specification, or number of partitions that the input space should have.

Change-Detection Using Contextual Information and Fuzzy Entropy Principle

This paper presents an unsupervised change detection method for computing the amount of changes that have occurred within an area by using remotely sensed technologies and fuzzy modeling. The discussion concentrates on the formulation of a standard procedure that, using the concept of fuzzy sets and fuzzy logic, can define the likelihood of changes detected from remotely sensed data. The fuzzy visualization of areas undergoing changes can be incorporated into a decision support system for prioritization of areas requiring environmental monitoring. One of the main problems related to unsupervised change detection methods lies in the lack of efficient automatic techniques for discriminating between changed and unchanged pixels in the difference image. Such discrimination is usually performed by using empirical strategies or manual trial-and-error procedures, which affect both, the accuracy and the reliability of the change-detection process. To overcome such drawbacks, in this paper, we propose an automatic technique for the analysis of the difference image. Such technique allows the automatic selection of the decision threshold. We used a thresholding approach by performing fuzzy partition on a two-dimensional (2-D) histogram, which included contextual information, based on fuzzy relation and maximum fuzzy entropy principle. Experimental results confirm the effectiveness of proposed technique.

Um Modelo de Análise de Crédito para o Cartão BNDES utilizando um Sistema Neuro-Fuzzy NFHB

Este trabalho apresenta um modelo para análise de crédito para o Cartão BNDES utilizando um sistema Neuro-Fuzzy Hierárquico BSP que constitui um passo importante para uma aproximação do BNDES com as micro, pequenas e médias empresas. Foram selecionadas trezentas e dezoito solicitações de crédito retiradas aleatoriamente dos pedidos feitos ao BNDES. Estas solicitações foram analisadas por especialistas em crédito, dos bancos emissores, que decidiram por conceder ou não o crédito solicitado. Após aplicado o modelo, os resultados alcançados, pelo classificador, foram considerados plenamente satisfatórios pelos especialistas, pois na totalidade dos casos que foram apresentados para serem avaliados pelo indutor, o modelo apresentou a mesma opinião do banco emissor.

The Hierarchical Neuro-Fuzzy BSP Model: An Application in Electric Load Forecasting

This work presents the development of a novel hybrid system called Hierarchical Neuro-Fuzzy BSP (HNFB) and its application in electric load forecasting. The HNFB system is based on the BSP partitioning (Binary Space Partitioning) of the input space and has been developed in order to bypass the traditional drawbacks of neuro-fuzzy systems: the reduced number of allowed inputs and the poor capacity to create their own structure. To test the HNFB system, we have used monthly load data of six electric energy companies. The results are compared with other forecast methods, such as Neural Networks and Box & Jenkins.