Selwyn Piramuthu

Financial credit-risk evaluation with neural and neurofuzzy systems

Credit-risk evaluation decisions are important for the financial institutions involved due to the high level of risk associated with wrong decisions. The process of making credit-risk evaluation decision is complex and unstructured. Neural networks are known to perform reasonably well compared to alternate methods for this problem. However, a drawback of using neural networks for credit-risk evaluation decision is that once a decision is made, it is extremely difficult to explain the rationale behind that decision. Researchers have developed methods using neural network to extract rules, which are then used to explain the reasoning behind a given neural network output. These rules do not capture the learned knowledge well enough. Neurofuzzy systems have been recently developed utilizing the desirable properties of both fuzzy systems as well as neural networks. These neurofuzzy systems can be used to develop fuzzy rules naturally. In this study, we analyze the beneficial aspects of using both neurofuzzy systems as well as neural networks for credit-risk evaluation decisions.

Identifying RFID-embedded objects in pervasive healthcare applications

The organization and delivery of pervasive healthcare have benefited much from advances in wireless systems. While wireless systems and their components have certainly enhanced the quality of pervasive healthcare administered in remote locations, their potential in other areas of healthcare cannot be underestimated. We consider Radio Frequency Identification (RFID) tags, which are increasingly being used in pervasive healthcare applications. Specifically, we study the dynamics of locating and identifying the presence of a tag in such systems. Although a tag may be present, it may not necessarily be visible to the tag reader due to various constraints or reasons. We propose and illustrate several algorithms for locating the presence of RFID tagged objects in the field of the reader and study their dynamics as well as their strengths and benefits. Our results indicate that the location accuracy of RFID tag readers can be improved through appropriate data collection as well as algorithms used for data inference.

Evaluating feature selection methods for learning in data mining applications

Abstract Recent advances in computing technology in terms of speed, cost, as well as access to tremendous amounts of computing power and the ability to process huge amounts of data in reasonable time has spurred increased interest in data mining applications to extract useful knowledge from data. Machine learning has been one of the methods used in most of these data mining applications. It is widely acknowledged that about 80% of the resources in a majority of data mining applications are spent on cleaning and preprocessing the data. However, there have been relatively few studies on preprocessing data used as input in these data mining systems. In this study, we evaluate several inter-class as well as probabilistic distance-based feature selection methods as to their effectiveness in preprocessing input data for inducing decision trees. We use real-world data to evaluate these feature selection methods. Results from this study show that inter-class distance measures result in better performance compared to probabilistic measures, in general.

Protocols for RFID tag/reader authentication

Radio-Frequency Identification (RFID) tags are poised to supplant barcodes in the very near future. Their information storage capacity as well as their ability to transfer information through contactless means without line-of-sight translates to significant advantage over barcodes. However, cost and privacy issues are major impediments to their widespread use. We consider the latter issue, specifically those that relate to securely authenticating RFID tags and readers. Light-weight authentication protocols are necessary in RFID applications due to tag-level constraints. Over the past few years, several streams of research have emerged approaching the RFID tag/reader privacy/security problem from different perspectives. We study and evaluate a few protocols from each of those streams, identify possible vulnerabilities, and provide alternate solutions when possible. We provide security analysis of the proposed solutions.

JOB SHOP SCHEDULING WITH A GENETIC ALGORITHM AND MACHINE LEARNING

Dynamic job shop scheduling has been proven to be an intractable problem for analytical procedures. Recent advances in computing technology, especially in artificial intelligence, have alleviated this problem by intelligently restricting the search space considered, thus opening the possibility of obtaining better results. Researchers have used various techniques that were developed under the general rubric of artificial intelligence to solve job shop scheduling problems. The most common of these have been expert systems, genetic algorithms and machine learning. Of these, we identify machine learning and genetic algorithms to be promising for scheduling applications in a job shop. In this paper, we propose to combine complementarily the strengths of genetic algorithms and induced decision trees, a machine learning technique, to develop a job shop scheduling system. Empirical results, using machine learning for releasing jobs into the shop floor and a genetic algorithm to dispatch jobs at each machine, are...

Knowledge-based framework for automated dynamic supply chain configuration

Supply chain management has gained renewed interest among researchers in recent years. This is primarily due to the availability of timely information across the various stages of the supply chain, and therefore the need to effectively utilize the information for improved performance. Although information plays a major role in effective functioning of supply chains, there is a paucity of studies that deal specifically with the dynamics of supply chains and how data collected in these systems can be used to improve their performance. In this paper I develop a framework, with machine learning, for automated supply chain configuration. Supply chain configuration used to be mostly a one-shot problem. Once a supply chain is configured, researchers and practitioners were more interested in means to improve performance given that initial configuration. However, recent developments in e-commerce applications and faster communication over the Internet in general necessitates dynamic (re)configuration of supply chains over time to take advantage of better configurations. Using examples, I show performance improvements of the proposed adaptive supply chain configuration framework over static configurations.

Framework for efficient feature selection in genetic algorithm based data mining

We present the design of more effective and efficient genetic algorithm based data mining techniques that use the concepts of feature selection. Explicit feature selection is traditionally done as a wrapper approach where every candidate feature subset is evaluated by executing the data mining algorithm on that subset. In this article we present a GA for doing both the tasks of mining and feature selection simultaneously by evolving a binary code along side the chromosome structure used for evolving the rules. We then present a wrapper approach to feature selection based on Hausdorff distance measure. Results from applying the above techniques to a real world data mining problem show that combining both the feature selection methods provides the best performance in terms of prediction accuracy and computational efficiency.

Challenges associated with RFID tag implementations in supply chains

Radio-Frequency Identification (RFID) tags are gaining widespread popularity throughout the supply chain from raw material acquisition, manufacturing, transportation, warehousing, retailing to the ultimate consumers. A majority of extant literature in this area explore the beneficial aspects of RFID tags such as their batch readability, resistance to harsh environmental conditions, information storage and processing capability, among others. Given the recent explosion of interest in RFID tag incorporation in supply chains, literature in the area has not yet comprehensively identified nor addressed associated challenges and impediments to successful implementations. We purport to fill this gap and to raise awareness by identifying and discussing critical issues such as ownership transfer, privacy/security, computing bottleneck, read error, and cost-benefit issues such as opportunity cost, risk of obsolescence, information sharing, and inter-operability standards.

A classification approach using multi-layered neural networks

Abstract There has been an increasing interest in the applicability of neural networks in disparate domains. In this paper, we describe the use of multi-layered perceptrons, a type of neural-network topology, for financial classification problems, with promising results. Back-propagation, which is the learning algorithm most often used in multi-layered perceptrons, however, is inherently an inefficient search procedure. We present improved procedures which have much better convergence properties. Using several financial classification applications as examples, we show the efficacy of using multilayered perceptrons with improved learning algorithms. The modified learning algorithms have better performance, in terms of classification/prediction accuracies, than the methods previously used in the literature, such as probit analysis and similarity-based learning techniques.

RFID-generated traceability for contaminated product recall in perishable food supply networks

As perishable food supply networks become more complex, incidents of contamination in these supply networks have become fairly common. Added to this complexity is the fact that there have been long delays in identifying the contamination source in several such incidents. Even when the contamination source was identified, there have been cases where the ultimate destination of all contaminated products were not known with complete certainty due, in part, to dispersion in these supply networks. We study the recall dynamics in a three-stage perishable food supply network through three different visibility levels in the presence of contamination. Specifically, we consider allocation of liability among the different players in the perishable supply network based on the accuracy with which the contamination source is identified. We illustrate the significance of finer levels of granularity both upstream and downstream as well as determine appropriate visibility levels and recall policies.