Shing Chiang Tan

Application of an adaptive neural network with symbolic rule extraction to fault detection and diagnosis in a power generation plant

Artificial neural networks have a good potential to be employed for fault diagnosis and condition monitoring problems in complex processes. In this paper, the applicability of the fuzzy ARTMAP (FAM) neural network as an intelligent learning system for fault detection and diagnosis in a power generation plant is described. The process under scrutiny is the circulating water (CW) system, with specific attention to the conditions of heat transfer and tube blockage in the CW system. A series of experiments has been conducted systematically to investigate the effectiveness of FAM in fault detection and diagnosis tasks. In addition, a set of domain rules has been extracted from the trained FAM network so that its predictions can be explained and justified. The outcomes demonstrate the benefits of employing FAM as an intelligent fault detection and diagnosis tool with an explanatory capability for monitoring and diagnosing complex processes in power generation plants.

Fuzzy ARTMAP dynamic decay adjustment: An improved fuzzy ARTMAP model with a conflict resolving facility

This paper presents a hybrid neural network classifier of fuzzy ARTMAP (FAM) and the dynamic decay adjustment (DDA) algorithm. The proposed FAMDDA model is a conflict-resolving classifier that can perform stable and incremental learning while settling overlapping of hyper-rectangular prototypes of different classes in minimizing misclassification rates. The performance of FAMDDA is evaluated using a number of benchmark data sets. The results are analyzed and compared with those from FAM and a number of machine learning classifiers. The outcomes show that FAMDDA has a better generalization capability than FAM, and its performance is comparable with those from other classifiers. The effectiveness of FAMDDA is also demonstrated in an application pertaining to condition monitoring of a circulating water system in a power generation station. Implications on the effectiveness of FAMDDA from the application point of view are discussed.

Fuzzy ARTMAP and hybrid evolutionary programming for pattern classification

In this paper, an Evolutionary Artificial Neural Network (EANN) that combines the Fuzzy ARTMAP (FAM) network and a Hybrid Evolutionary Programming (HEP) model is introduced. The proposed FAM-HEP model, which combines the strengths of FAM and HEP, is able to construct its network structure autonomously as well as to perform learning and evolutionary search and adaptation concurrently. The effectiveness of the proposed FAM-HEP network is assessed empirically using several benchmark data sets and a real medical diagnosis problem. The performance of FAM-HEP is analyzed, and the results are compared with those of FAM-EP, FAM, and other classification models. In general, the results of FAM-HEP are better than those of FAM-EP and FAM, and are comparable with those from other classification models. The study also reveals the potential of FAM-HEP as an innovative EANN model for undertaking pattern classification problems in general, and a promising computerized decision support tool for tackling medical diagnosis tasks in particular.

A Truly Online Learning Algorithm using Hybrid Fuzzy ARTMAP and Online Extreme Learning Machine for Pattern Classification

This paper presents a Hybrid Fuzzy ARTMAP (FAM) and Online Extreme learning machine (OELM), hereafter denoted as FAM-OELM, which enables online learning to start from the first trained data samples without having to set up an initialization phase which requires a chunk of data samples to be ready prior to training. The idea of developing FAM-OELM is motivated by the ELM concept proposed by Huang et al., for being an efficient learning algorithm that provides better generalization performance at a much faster learning speed. However, different from the batch learning ELM and its variant called the online sequential extreme learning machine which still requires an initial offline training phase before it can turn into online training, the proposed FAM-OELM showcases a framework that enable online learning to commence right from the first data sample. Here, classification can be conducted at any time during the training phase. Such appealing feature of the proposed algorithm has strictly fulfilled the criteria of being truly sequential, while many of the existing algorithms are not. In addition, FAM-OELM automatically grows hidden neuron such that the network can accommodate new information without over fitting and compromising on the knowledge learnt earlier. The simulation results reveal the efficacy and validity of FAM-OELM when it is applied to a real world application and various benchmark problems.

On the reduction of complexity in the architecture of fuzzy ARTMAP with dynamic decay adjustment

Abstract This paper presents a hybrid network (FAMDDA-T) comprising the Fuzzy ARTMAP (FAM) neural network and the Dynamic Decay Adjustment (DDA) algorithm with an online pruning strategy. Twelve benchmark datasets are used to demonstrate the effectiveness of FAMDDA-T. The results of FAMDDA-T are compared with those of FAMDDA (without pruning), and the Radial Basis Function Network with DDA (RBFN-DDA) as well as its pruning version (RBFN-DDA-T). It is observed that, when compared with other DDA-based networks, FAMDDA-T is able to form a parsimonious network structure and, at the same time, to maintain a high level of network generalization in tackling pattern classification problems.

A hybrid FAM---CART model and its application to medical data classification

In this paper, a hybrid model consisting of the fuzzy ARTMAP (FAM) neural network and the classification and regression tree (CART) is formulated. FAM is useful for tackling the stability–plasticity dilemma pertaining to data-based learning systems, while CART is useful for depicting its learned knowledge explicitly in a tree structure. By combining the benefits of both models, FAM–CART is capable of learning data samples stably and, at the same time, explaining its predictions with a set of decision rules. In other words, FAM–CART possesses two important properties of an intelligent system, i.e., learning in a stable manner (by overcoming the stability–plasticity dilemma) and extracting useful explanatory rules (by overcoming the opaqueness issue). To evaluate the usefulness of FAM–CART, six benchmark medical data sets from the UCI repository of machine learning and a real-world medical data classification problem are used for evaluation. For performance comparison, a number of performance metrics which include accuracy, specificity, sensitivity, and the area under the receiver operation characteristic curve are computed. The results are quantified with statistical indicators and compared with those reported in the literature. The outcomes positively indicate that FAM–CART is effective for undertaking data classification tasks. In addition to producing good results, it provides justifications of the predictions in the form of a decision tree so that domain users can easily understand the predictions, therefore making it a useful decision support tool.

Evolutionary Fuzzy ARTMAP Neural Networks and their Applications to Fault Detection and Diagnosis

In this paper, two mutation-based evolving artificial neural networks, which are based on the Fuzzy ARTMAP (FAM) network and evolutionary programming, are proposed. The networks utilize the knowledge base extracted from a set of data to perform search and adaptation. The performances of the two networks are assessed using benchmark problems, with the results analyzed and discussed. The effects of the network parameters are evaluated through a parametric study. The applicability of the networks is also demonstrated using a real fault detection and diagnosis task in a power generation plant. The experimental results consistently indicate the usefulness of the proposed evolutionary FAM-based networks in yielding good classification performances with parsimonious network structures.

An online pruning strategy for supervised ARTMAP-based neural networks

Identifying an appropriate architecture of an artificial neural network (ANN) for a given task is important because learning and generalisation of an ANN is affected by its structure. In this paper, an online pruning strategy is proposed to participate in the learning process of two constructive networks, i.e. fuzzy ARTMAP (FAM) and fuzzy ARTMAP with dynamic decay adjustment (FAMDDA), and the resulting hybrid networks are called FAM/FAMDDA with temporary nodes (i.e. FAM-T and FAMDDA-T, respectively). FAM-T and FAMDDA-T possess a capability of reducing the network complexity online by removing unrepresentative neurons. The performances of FAM-T and FAMDDA-T are evaluated and compared with those of FAM and FAMDDA using a total of 13 benchmark data sets. To demonstrate the applicability of FAM-T and FAMDDA-T, a real fault detection and diagnosis task in a power plant is tested. The results from both benchmark studies and real-world application show that FAMDDA-T and FAM-T are able to yield satisfactory classification performances, with the advantage of having parsimonious network structures.

A hybrid neural network classifier combining ordered fuzzy ARTMAP and the dynamic decay adjustment algorithm

This paper presents a novel conflict-resolving neural network classifier that combines the ordering algorithm, fuzzy ARTMAP (FAM), and the dynamic decay adjustment (DDA) algorithm, into a unified framework. The hybrid classifier, known as Ordered FAMDDA, applies the DDA algorithm to overcome the limitations of FAM and ordered FAM in achieving a good generalization/performance. Prior to network learning, the ordering algorithm is first used to identify a fixed order of training patterns. The main aim is to reduce and/or avoid the formation of overlapping prototypes of different classes in FAM during learning. However, the effectiveness of the ordering algorithm in resolving overlapping prototypes of different classes is compromised when dealing with complex datasets. Ordered FAMDDA not only is able to determine a fixed order of training patterns for yielding good generalization, but also is able to reduce/resolve overlapping regions of different classes in the feature space for minimizing misclassification during the network learning phase. To illustrate the effectiveness of Ordered FAMDDA, a total of ten benchmark datasets are experimented. The results are analyzed and compared with those from FAM and Ordered FAM. The outcomes demonstrate that Ordered FAMDDA, in general, outperforms FAM and Ordered FAM in tackling pattern classification problems.

Introduction to Evolutionary Computing in System Design

In this chapter, an introduction on the use of evolutionary computing techniques, which are considered as global optimization and search techniques inspired from biological evolutions, in the domain of system design is presented. A variety of evolutionary computing techniques are first explained, and the motivations of using evolutionary computing techniques in tackling system design tasks are then discussed. In addition, a number of successful applications of evolutionary computing to system design tasks are described.