International Journal of New Computer Architectures and their Applications

810-820

Authors: Mohd Ibrahim Shapiai, Zuwairie Ibrahim, Marzuki Khalid, Lee Wen Jau, Soon-Chuan Ong, Vladimir Pavlovich

The cost of the experimental setup during the assembly process development of a chipset, particularly the under-fill process, can often result in insufficient data samples. In INTEL Malaysia, for example, the historical chipset data from an under-fill process consist of only a few samples. As a result, existing machine learning algorithms for predictive modeling cannot be applied in this setting. Despite this challenge, the use of data driven decisions remains critical for further optimization of this engineering process. In the proposed framework, the original weighted kernel regression with correlation factor (WKRCF) is strengthened by normalizing the input parameters and employing the Particle Swarm Optimization (PSO) as weight estimator. It is found that PSO gives flexibility in defining the objective function as compared to the iteration technique of WKRCF. Thus, an assumption on noise contamination to the available training samples can be implemented. Even though only four samples are used during the training stage of the conducted experiment, the proposed approach is able to provide better prediction within the engineer’s requirements as compared with WKRCF. Thus, the proposed approach is beneficial for recipe generation in an assembly process development.

838-847

Authors: M. Z. Rehman, N. M. Nawi

The traditional Back-propagation Neural Network (BPNN) Algorithm is widely used in solving many real time problems in world. But BPNN possesses a problem of slow convergence and convergence to local minima. Previously, several modifications are suggested to improve the convergence rate of Gradient Descent Back-propagation algorithm such as careful selection of initial weights and biases, learning rate, momentum, network topology, activation function and ‘gain’ value in the activation function. This research proposed an algorithm for improving the current working performance of Back-propagation algorithm by adaptively changing the momentum value and at the same time keeping the ‘gain’ parameter fixed for all nodes in the neural network. The performance of the proposed method known as ‘Gradient Descent Method with Adaptive Momentum (GDAM)’ is compared with the performances of ‘Gradient Descent Method with Adaptive Gain (GDM-AG)’ and ‘Gradient Descent with Simple Momentum (GDM)’. The efficiency of the proposed method is demonstrated by simulations on five classification problems. Results show that GDAM can be used as an alternative approach for BPNN because it demonstrate better accuracy ratio on the chosen classification problems.

866-878

Authors: Norhamreeza Abdul Hamid, Nazri Mohd Nawi, Rozaida Ghazali, Mohd Najib Mohd Salleh

In some practical Neural Network (NN) applications, fast response to external events within enormously short time is highly demanded. However, by using back propagation (BP) based on gradient descent optimisation method obviously not satisfy in several application due to serious problems associated with BP which are slow learning convergence velocity and confinement to shallow minima. Over the years, many improvements and modifications of the BP learning algorithm have been reported. In this research, we modified existing BP learning algorithm with adaptive gain by adaptively change the momentum coefficient and learning rate. In learning the patterns, the simulation results indicate that the proposed algorithm can hasten up the convergence behaviour as well as slide the network through shallow local minima compare to conventional BP algorithm. We use five common benchmark classification problems to illustrate the improvement of the proposed algorithm.

894-910

Authors: Lustiana Pratiwi, Yun-Huoy Choo, Azah Kamilah Muda

This paper proposed to generate solution for Particle Swarm Optimization (PSO) algorithms using Ant Colony Optimization approach, which will satisfy the Gaussian distributions to enhance PSO performance. Coexistence, cooperation, and individual contribution to food searching by a particle (ant) as a swarm (ant) survival behavior, depict the common characteristics of both algorithms. Solution vector of ACO is presented by implementing density and distribution function to search for a better solution and to specify a probability functions for every particle (ant). Applying a simple pheromone-guided mechanism of ACO as local search is to handle P ants equal to the number of particles in PSO and generate components of solution vector, which satisfy Gaussian distributions. To describe relative probability of different random variables, PDF and CDF are capable to specify its own characterization of Gaussian distributions. The comparison is based on the experimental result to increase higher fitness value and gain better reducts, which has shown that PDF is better than CDF in terms of generating smaller number of reducts, improved fitness value, lower number of iterations, and higher classification accuracy.

926-941

Authors: Ben Sadeh, Sundar Gopalakrishnan

Unit testing is a method for quickly assessing the building blocks of a program and for obtaining accurate error localizations. However, in order to achieve these qualities, the tests cases need to be isolated, since an external call may imply a connection to a remote database. This requirement also makes unit testing difficult to initiate for classes with outside dependencies, and consequently several approaches have been devised to facilitate unit testing of these methods. This paper focuses on the different ways of unit testing Java methods with external dependencies in an Android application. Additionally, it covers a new category of testing methodology called shadow objects. First, the study examines some of the current methods of testing. Then, it details the different ways a class with an external dependency could be unit tested. Finally, the paper presents a discussion and evaluation of the study.

953-976

Authors: Mohammed Wajahat Kamal, Moataz A. Ahmed

Reliable and accurate software development effort estimation has always been a daunting task for project managers. More recently, the use of Use Cases for software effort estimation has gained wide popularity. Researchers from academia as well as industry have shown interest in the Use Case based approaches because of the promising results obtained along with their early applicability. There has been a number of approaches proposed in the literature. However, there is no criteria that could be used to aid practitioners in selecting appropriate approaches suitable for their particular development efforts. In this paper we present a set of attribute-based criteria to classify and compare these approaches and provide such aid to practitioners. The set is also meant to guide researchers interested in proposing new use case-based approaches. The paper conducts a systematic review of a number of representative Use Case-based effort estimation approaches against the criteria. Analysis of the discussion highlights some open issues for future research. Addressing some of the issues, we present and discuss a framework for developing use case-based effort estimation models.

991- 1000

Authors: Elfadil Sabeil, Azizah Bt Abdul Manaf, Zuraini Ismail, Mohamed Abas

Lack of Cyber Forensics experts is a huge challenge facing the world today. It comes due to the fancy of Cyber Forensics training or education. The multidisciplinary nature of Cyber Forensics proliferates to diverse training programmes, from a handful day‟s workshop to Postgraduate in Cyber Forensics. Consequently, this paper concentrates on analyzing the Cyber Forensics training programmes in terms of Competency-Based Framework. The study proves that Cyber Forensics training or education has improper Competency-Based Framework.