Abdul Razak Hamdan

Multi-population cooperative bat algorithm-based optimization of artificial neural network model

The performance of an artificial neural network (ANN) depends on the connection weights and network structure. Many optimization algorithms have been applied for ANN model selection. This paper presents an optimization algorithm based on the cooperative bat-inspired algorithm. The advantage of the bat algorithm lies in the combination of a population-based algorithm and local search; however, it is more powerful in local search. Therefore to better balance exploration and exploitation in the population some modifications to the velocity equation of the standard bat algorithm are applied. In addition, we propose two new topologies for cooperation between subpopulations to further improve the algorithms capability to maintain the diversity of bats in the population. The first is a combination of two known mechanisms (Ring and Master-Slave), and the second inserts a Coevolving strategy of slave subpopulations in the Master-Slave strategy. The proposed methods are applied for the selection of an ANN model, where both the structure of the ANN and its weights are optimized by the method. Six classification and two time series prediction benchmark datasets are tested and the performance of the proposed algorithms is evaluated and compared with other methods in the literature. Statistical analysis shows that for the classification problem there is a significant improvement in the bat algorithm with Ring and Master-Slave strategies cooperation compared to the other methods in the literature in terms of classification error for three cases out of five and a significant enhancement in the number of connection weights in the network. The analysis also shows that for time series prediction there is a significant improvement in the prediction error for all the cases.

Optimization of neural network model using modified bat-inspired algorithm

A modified bat algorithm with a new solution representation for both optimizing the weights and structure of ANNs is proposed.To improve the exploration and exploitation capability of bat algorithm some modifications based on chaotic map on bat algorithm are studied.The Taguchi method is used to tune the parameters of the algorithm.Six classifications and two time series benchmark datasets are used to test the performance of the proposed approach in terms of classification and prediction accuracy.Finally, our best method is applied to a real-world problem, namely to predict the future values of rainfall data in Selangor at Malaysia. The success of an artificial neural network (ANN) strongly depends on the variety of the connection weights and the network structure. Among many methods used in the literature to accurately select the network weights or structure in isolate; a few researchers have attempted to select both the weights and structure of ANN automatically by using metaheuristic algorithms. This paper proposes modified bat algorithm with a new solution representation for both optimizing the weights and structure of ANNs. The algorithm, which is based on the echolocation behaviour of bats, combines the advantages of population-based and local search algorithms. In this work, ability of the basic bat algorithm and some modified versions which are based on the consideration of the personal best solution in the velocity adjustment, the mean of personal best and global best solutions through velocity adjustment and the employment of three chaotic maps are investigated. These modifications are aimed to improve the exploration and exploitation capability of bat algorithm. Different versions of the proposed bat algorithm are incorporated to handle the selection of the structure as well as weights and biases of the ANN during the training process. We then use the Taguchi method to tune the parameters of the algorithm that demonstrates the best ability compared to the other versions. Six classifications and two time series benchmark datasets are used to test the performance of the proposed approach in terms of classification and prediction accuracy. Statistical tests demonstrate that the proposed method generates some of the best results in comparison with the latest methods in the literature. Finally, our best method is applied to a real-world problem, namely to predict the future values of rainfall data and the results show satisfactory of the method.

Hybrid feature selection based on enhanced genetic algorithm for text categorization

An enhanced genetic algorithm (EGA) is proposed to reduce text dimensionality.The proposed EGA outperformed the traditional genetic algorithm.The EGA is incorporated with six filter feature selection methods to create hybrid feature selection approaches.The proposed hybrid approaches outperformed the single filtering methods. This paper proposes hybrid feature selection approaches based on the Genetic Algorithm (GA). This approach uses a hybrid search technique that combines the advantages of filter feature selection methods with an enhanced GA (EGA) in a wrapper approach to handle the high dimensionality of the feature space and improve categorization performance simultaneously. First, we propose EGA by improving the crossover and mutation operators. The crossover operation is performed based on chromosome (feature subset) partitioning with term and document frequencies of chromosome entries (features), while the mutation is performed based on the classifier performance of the original parents and feature importance. Thus, the crossover and mutation operations are performed based on useful information instead of using probability and random selection. Second, we incorporate six well-known filter feature selection methods with the EGA to create hybrid feature selection approaches. In the hybrid approach, the EGA is applied to several feature subsets of different sizes, which are ranked in decreasing order based on their importance, and dimension reduction is carried out. The EGA operations are applied to the most important features that had the higher ranks. The effectiveness of the proposed approach is evaluated by using naive Bayes and associative classification on three different collections of Arabic text datasets. The experimental results show the superiority of EGA over GA, comparisons of GA with EGA showed that the latter achieved better results in terms of dimensionality reduction, time and categorization performance. Furthermore, six proposed hybrid FS approaches consisting of a filter method and the EGA are applied to various feature subsets. The results showed that these hybrid approaches are more effective than single filter methods for dimensionality reduction because they were able to produce a higher reduction rate without loss of categorization precision in most situations.

Solving a practical examination timetabling problem: a case study

This paper presents a Greedy-Least Saturation Degree (G-LSD) heuristic (which is an adaptation of the least saturation degree heuristic) to solve a real-world examination timetabling problem at the University Kebangsaan, Malaysia. We utilise a new objective function that was proposed in our previous work to evaluate the quality of the timetable. The objective function considers both timeslots and days in assigning exams to timeslots, where higher priority is given to minimise students having consecutive exams on the same day. The objective also tries to spread exams throughout the examination period. This heuristic has the potential to be used for the benchmark examination datasets (e.g. the Carter datasets) as well as other real world problems. Computational results are presented.

Employability and service science: Facing the challenges via curriculum design and restructuring

A university faces many challenges in equipping graduates with the right skills and attitudes that will make them employable. The shift from the predominantly manufacturing based industries to the service based one is a source of one such challenge. Focusing on the Faculty of Information Science and Technology, it is contended that there is a need to examine its current curriculum and methods of instruction so as to ensure that future graduates will be suitable for this new economy. In this position paper, based on the model of the T-shaped professional and drawing insights from the newly developed field of Service Science, Management and Engineering (SSME), it is proposed that a combination of approaches is undertaken in restructuring the curriculum so that future graduates from the faculty in particular, and the university in general, will be ready to face the challenges posed by the new era of services.

Improved Dynamic Ant Colony System (DACS) on symmetric Traveling Salesman Problem (TSP)

Ants are a fascinating creature that demonstrates a capability of finding food and bring it back to their nest. Their ability as a colony to find paths or routes to the food sources has inspired the development of an algorithm namely ant colony system (ACS). The principle of cooperation has been the backbone in these algorithmic developments. However, observing the behavior of a single ant can provide an added value to the principle. Ants communicate to each other through a chemical substance called pheromone. Manipulating and empowering this substance is the trivial factor in finding the best solution. However, without considering the experiences of individuals would contribute a complete waste of available knowledge. Having the concepts of a single ant trying to reconstruct or reconnect the paths that was previously laid by its colony when a certain obstacle placed on its normal paths has added another level of pheromone updates. Thus, this new level of pheromone updates which manipulating and empowering the searching experiences of individual ants can improve the current ACS algorithm. Traveling salesman problem (TSP) was used as a case study to show the capability of the algorithm in order to find the best solution in terms of the shortest distance. At the end of this paper, we presented an experimental result on a benchmark data to show how it could improve the fundamental of ACS algorithm.

Development of Intelligent Visual Inspection System (IVIS) for Bottling Machine

This paper presents a research on developing an intelligent visual inspection system (IVIS) for bottling machine, focusing on the development of image processing framework for defect detection. The objective of the research is to contribute a method on modeling, integrating and enhancing IVIS for the process of quality control in industrial area. IVIS application for quality control was studied using plastic bottles on a production line simulation. An experiment had done by using developed software and special equipments such as conveyor belt, lighting source, and a Web camera (Webcam) to capture the image. The experiment result shows that the system is accurate enough to detect moving object on the speed at 106 rpm with the accuracy of the image acquisition is 94.264%.

Classification for talent management using Decision Tree Induction techniques

Classification is one of the tasks in Data mining. Nowadays, there are many classification techniques being used to solve classification problems such as Neural Network, Genetic Algorithm, Bayesian and others. In this article, we attempt to present a study on how talent management can be implemented using Decision Tree Induction techniques. By using this approach, talent performance can be predicted using past experience knowledge discovered from the existing database. In the experimental phase, we use selected classification algorithms from Decision tree techniques to propose suitable classifier for the dataset. As a result, the C4.5 classifier algorithm shows the highest accuracy of model for the dataset. Consequently, the possible talent rules are generated based on C4.5 classifier especially for the talent forecasting purposes.

Outlier detection based on rough sets theory

An outlier in a dataset is a point or a class of points that is considerably dissimilar to or inconsistent with the remainder of the data. Detection of outliers is important for many applications and has always attracted attention among data mining research community. In this paper, a new method in detecting outlier based on Rough Sets Theory is proposed. The main concept of using the Rough Sets for outlier detection is to discover Non-Reduct from the information system (IS). Non-Reduct is a set of attributes from IS that may contain outliers. It is discovered through the computation of Non-Reduct by defining Indiscernibility matrix modulo (iDMM D) and Indiscernibility function modulo (iDFM D). A measurement called RSetOF (Rough Set Outlier Factor Value) is hereby defined to identify and detect outlier objects. Extensive experiments were conducted where ten benchmark datasets were tested with the proposed method. To evaluate the effectiveness of performance of the proposed method, RSetAlg is compared to the Frequent Pattern (FindFPOF) method. The experimental result reveals that the approach utilised is a good outlier detection method compared to FindFPOF method. Thus, this proposed method has formed a novel and competitive method in outlier detection.

Potential intelligent techniques in Human Resource Decision Support System (HR DSS)

Human resource decisions are subject to limitations, because they always depend on human knowledge, judgement and preference. Decision support applications can be used to provide fair and consistent decisions, besides to improve the effectiveness of decision making processes. An intelligent decision support system (IDSS) is developed to assist decision makers in high level phases of decision making by integrating human knowledge with modeling tools. In this paper, we described the potential to implement the IDSS in human resource management (HRM) activities. This study consists of three parts; the first part is to understand the IDSS concepts, applications and related research in human resources decision making application known as HR DSS. The second part is to identify the potential intelligent techniques that can be used in HR DSS application, and the third part is to suggest the HR DSS framework that is related to human resource decisions. Finally, the paper proposed the HR DSS framework and the potential intelligent techniques that can be used to develop the IDSS application in any phases of decision making processes.