Norita Md Norwawi

A review of feature selection techniques via gene expression profiles

The invention of DNA microarray technology has modernized the approach of biology research in such a way that scientists can now measure the expression levels of thousands of genes simultaneously in a single experiment. Although this technology has shifted a new era in molecular classification, interpreting microarray data still remain a challenging issue due to their innate nature of “high dimensional low sample size”. Therefore, robust and accurate feature selection methods are required to identify differentially expressed genes across varied samples for example between cancerous and normal cells. Successful of feature selection techniques will assist to correctly classify different cancer types and consequently led to a better understanding of genetic signatures in cancers and would improve treatment strategies. This paper presents a review of feature selection techniques that have been employed in microarray data analysis. Moreover, other problems associated with microarray data analysis also addressed. In addition, several trends were noted including highly reliance on filter techniques compared to wrapper and embedded, a growing direction towards ensemble feature selection techniques and future extension to apply feature selection in combination of heterogeneous data sources.

Modelling of Reservoir Water Release Decision Using Neural Network and Temporal Pattern of Reservoir Water Level

The reservoir is one of flood mitigation methods that aim to reduce the effect of flood at downstream flood prone areas. At the same time the reservoir also serves other purposes. Through modelling, how the reservoir operator made decisions in the past can be revealed. Consequently, the information can be used to guide reservoir operator making present decision especially during emergency situations such as flood and drought. This paper discussed modelling of reservoir water release decision using Neural Network (NN) and the temporal pattern of reservoir water level. Temporal pattern is used to represent the time delay as the rainfall upstream may not directly raise the reservoir water level. The flow of water may take some time to reach the reservoir due to the location. Seven NN models have been developed and tested. The findings show that the NN model with 5-25-1 architecture demonstrate the best performance compare to the other models.

Systematic review of web application security development model

In recent years, web security has been viewed in the context of securing the web application layer from attacks by unauthorized users. The vulnerabilities existing in the web application layer have been attributed either to using an inappropriate software development model to guide the development process, or the use of a software development model that does not consider security as a key factor. Therefore, this systematic literature review is conducted to investigate the various security development models used to secure the web application layer, the security approaches or techniques used in the process, the stages in the development model in which the approaches or techniques are emphasized, and the tools and mechanism used to detect vulnerabilities. The study extracted 499 publications from respectable scientific sources, i.e. the IEEE Computer Society, ACM Digital Library, Google-Scholar, Science Direct, Scopus, Springer Link and ISI Web. After investigation, only 43 key primary studies were considered for this review based on defined inclusion and exclusion criteria. From the review, it appears that no one development model is referred to as a standard or preferred model for web application development. However, agile development models seem to have gained more attention, probably due to the multiple stakeholders that are involved in discussing security viewpoints, rather than a few members of the development team. It appears also that there is consistency in the use of the threat-modeling technique, probably due to its effectiveness in dealing with different kinds of vulnerabilities.

Security metrics to improve misuse case model

Assessing security at an early stage of the web application development life cycle helps to design a secure system that can withstand malicious attacks. Measuring security at the requirement stage of the system development life cycle assists in mitigating vulnerabilities and increasing the security of the developed system, which reduces cost and rework. In this paper, we present a security metrics model based on the Goal Question Metric approach, focusing on the design of the misuse case model. The security metrics model assists in examining the misuse case model to discover and fix defects and vulnerabilities before moving to the next stages of system development. The presented security metrics are based on the OWASP top 10-2010, in addition to misuse case modelling antipattern.

Identifying student and organization matching pattern using Apriori algorithm for practicum placement

Students placement in industry for the practicum training is difficult due to the large number of students and organizations involved. Further the matching process is complex due to the various criteria set by the organization and students. This paper will discuss the results of a pattern extraction process using association rules of data mining technique where Apriori algorithm was chosen. The data use consists of Bachelor of Information Technology and Bachelor in Multimedia students of University Utara Malaysia from the year 2004 till 2005. Two experiments were conducted using undirected data and directed data. The pattern extracted gave information on the previous matching process done by University -Industry Linkage Centre of University Utara Malaysia.

Mining the student programming performance using rough set

One of the powerful data mining analysis is it can generates different set of knowledge when similar problem is presented to different data mining techniques. In this paper, a programming dataset was mined using rough set in order to investigate the significant factors that may influence students programming performance based on information from previous student performance. Then, the result was compared with other researches which had previously explored the data using statistic, clustering, and association rule. The dataset consists of 419 records with 70 attributes were pre-processed and then mined using rough set. The result indicates rough set has identified several new characteristics. The student who has been exposed to programming prior to entering university and obtained average score in Mathematics, English, and Malay Language subject during secondary Malaysian School Certificate (SPM) examination were among strong indicators that contributes to good programming grades. Besides that, the personality factor; the investigative and social type plus average cognitive person were also found as important factors that influence programming. This finding can be a guideline for the faculty to plan teaching and learning program for new registered student.

Hybrid of rough set theory and Artificial Immune Recognition System as a solution to decrease false alarm rate in intrusion detection system

Denial of Service (DoS) attacks is one of the security threats for computer systems and applications. It usually make use of software bugs to crash or freeze a service or network resource or bandwidth limits by making use of a flood attack to saturate all bandwidth. Predicting a potential DOS attacks would be very helpful for an IT departments or managements to optimize the security of intrusion detection system (IDS). Nowadays, false alarm rates and accuracy become the main subject to be addressed in measuring the effectiveness of IDS. Thus, the purpose of this work is to search the classifier that is capable to reduce the false alarm rates and increase the accuracy of the detection system. This study applied Artificial Immune System (AIS) in IDS. However, this study has been improved by using integration of rough set theory (RST) with Artificial Immune Recognition System 1 (AIRS1) algorithm, (Rough-AIRS1) to categorize the DoS samples. RST is expected to be able to reduce the redundant features from huge amount of data that is capable to increase the performance of the classification. Furthermore, AIS is an incremental learning approach that will minimize duplications of cases in a knowledge based. It will be efficient in terms of memory storage and searching for similarities in Intrusion Detection (IDS) attacks patterns. This study use NSL-KDD 20% train dataset to test the classifiers. Then, the performances are compared with single AIRS1 and J48 algorithm. Results from these experiments show that Rough-AIRS1 has lower number of false alarm rate compared to single AIRS but a little bit higher than J48. However, accuracy for this hybrid technique is slightly lower compared to others.

Extending V-model practices to support SRE to build secure web application

Technological advancements and rapid growth in the use of the Internet by the society have had a huge impact on information security. It has triggered the need for a major shift in the way web applications are developed. The high level security of these applications is crucial to their success. Therefore, information security has become a core requirement for producing trustworthy software driven by the need to guard critical assets. To develop a web application with adequate security features, it is highly recommended to capture security requirements early in the development lifecycle. In this paper, we propose a way of extending the V-Model requirements engineering phase to aid developers to engineer security requirements for a web application being developed, as well as, handling the security test planning. The aim is to support the proactive definition of security requirements by integrating security requirements engineering (SRE) activities with requirements engineering (RE) activities of the V-model.

Intelligent Decision Support Model Based on Neural Network to Support Reservoir Water Release Decision

Reservoir is one of the emergency environments that required fast an accurate decision to reduce flood risk during heavy rainfall and contain water during less rainfall. Typically, during heavy rainfall, the water level increase very fast, thus decision of the water release is timely and crucial task. In this paper, intelligent decision support model based on neural network (NN) is proposed. The proposed model consists of situation assessment, forecasting and decision models. Situation assessment utilized temporal data mining technique to extract relevant data and attribute from the reservoir operation record. The forecasting model utilize NN to perform forecasting of the reservoir water level, while in the decision model, NN is applied to perform classification of the current and changes of reservoir water level. The simulations have shown that the performances of NN for both forecasting and decision models are acceptably good.

Interacted Multiple Ant Colonies Optimization Approach to Enhance the Performance of Ant Colony Optimization Algorithms

One direction of ant colony optimization researches is dividing the ants’ population into several colonies. These colonies work together to collectively solve an optimization problem. This approach offers good opportunity to explore a large area of the search space. It seems to be a suitable approach to improve the performance of ant algorithms. This paper proposes a new generic algorithmic approach utilizing multiple ant colonies with some new interaction techniques. Computational test shows promising results of the new approach. The proposed approach outperforms the single colony ant algorithms in term of the solution quality with the same computational effort.