Software systems that are capable of controlling a network of computers for malicious intervention which focus on defraud and inspecting information are known as intrusion detection systems (IDS). Constantly changing and the complicated nature of intrusion activities on computer networks cannot be dealt with IDSs that are currently operational. In this paper, a Random Forests method based on the averaging method is proposed as a novel method to predict the types of intrusion attacks. Support vector classification model and nearest centroid classification model are used as the comparison models. The experimental results indicated that the intended model performed as well as the most advanced models like decision trees and outperforms the state of the art techniques like support vector classification models and nearest centroid classification model for the mentioned dataset with respect to parameters such as accuracy, the detection rate and false alarm.

Intrusion detection system, Random forests, Support vector classification, Accuracy, Detection rate, False Alarm.

The component-based software engineering (CBSE) follows the process of reusability and reconfiguration of components to achieve the better productivity. The context-aware systems are part of CBSE, which monitors the behavior of the system and adopt automatically according to the execution context. In this paper, we are concentrating on the aware context policies that automatically adapt to the given context given by the clients and redesign the software architecture based on the requirements. The component repository was introduced, where it contains the number of reusable components. The fuzzy logic was applied to the component selection in the component repository. The GRASP algorithm is used to optimize the system architecture. The dish TV middleware is used to test the adaptability of the system.

Components, GRASP, Fuzzy adapter, Context, Ontology.

It is difficult to maintain the route from source to destination in delay tolerant networks. The opportunistic routing overcomes the issues of route discovery in MANETs. The proposed routing protocol is a table driven approach for routing mechanism. It will effectively find the route between the source and destination by introducing the new concept called as hubs. The opportunistic adaptive routing protocol finds the fitness value of each node participating in the network and includes the nodes in the network which have the highest fitness value. The experimental results show the effectiveness of the algorithm.

Opportunistic routing, MANETs, Fitness, Hub, Nodes.

Assessment is an important procedure in teaching process at high schools. The assessment results may affect teaching programs and the methods. Teachers need to use assessment as a part of their teaching method. Hence, we propose a competence oriented innovation of assessing mathematics learning outcomes in Vietnam.

Teachers’ assessment competency, classroom assessment, mathematics teaching methods.

This study aims to investigate the effectiveness of GeoGebra package on learning outcomes of Mathematics students. There is necessity for this research due to the persistent failure recorded by the students in the country, especially at the external examinations like WAEC and NECO. Little or no improvement have been noticed despite numerous recommendations from the past researchers and there was indeed the need to incorporate GeoGebra, an ICT package, into the teaching and learning of Mathematics. This study adopted the non-equivalent pre-test post-test control group design. The study population comprised secondary school Mathematics students in Ogbomoso North L.G.A. of Oyo State, Nigeria. SS II Mathematics students from two intact classes from each of the two purposively selected schools in the area constituted the sample. The schools were selected on the basis of availability of functional computer systems. The classes were assigned into experimental and control groups using simple random sampling technique. The study concluded that the incorporation of GeoGebra and other ICT packages would improve the students’ learning outcomes in Mathematics, especially on students’ performance in both internal and external examinations; while their attitude towards Mathematics would also be positively enhanced.

GeoGebra, Millennium, Package, Mathematics, Instruction, Nigeria.

Regression is a criterion for estimating the relationships among variables. Though many authors have derived regression estimator on double sampling by making use of auxiliary variables without stratification and obtained their variances. Hence, there is need to determine a modified regression estimator on double sampling by making use of auxiliary variables with stratification for estimating the population estimates. This study proposed an estimator for estimating the population estimates of double sampling involving stratification. The data were collected on cost of House Rent, Transportation, Feeding, Fueling of Car, PHCN Bill, Toiletries, Fueling of Generator, Recharge Card and total expenditure of some randomly selected staff members of Federal University of Agriculture, Abeokuta, Ogun State, Nigeria. A random sample of size 138 was randomly selected from 256 respondents and sub sample of 81 was taken from the first sample. Auxiliary variables in the first sample and sub-sample (cost of House Rent, Transportation, Feeding, Fueling of Car; PHCN Bill, Toiletries, Fueling of Generator and Recharge Card) were stratified into two vector groups using level of expenditure and income as stratifying factors respectively. The mean and the minimum variance of the double sampling for the existing estimator are 92,634 and 2,672,766,891 respectively, while the mean and variance of the proposed estimator are 196,330.1 and 3,114,477 respectively. The minimum variance of the proposed estimator is smaller than the minimum variance of the existing estimator; hence, the proposed estimator is an efficient estimator.

Double sampling, Stratification, Auxiliary variables.

Healthcare insurance, delivered via the National Health Insurance Scheme (NHIS) is a veritable tool for making quality healthcare available to the majority of Nigerian citizens, irrespective of income status. Segmentation of dataset comprising of several features is a drawback of many applications. K-means clustering is a method used to cluster higher-dimensional dataset. This scheme, however, faces imminent collapse if an effective way of grouping health insurance claims is not found. Health insurance claims account for a significant portion of all claims received by insurers amounting to billions of naira annually. Thus, this study focused on application of data mining techniques that could help to drastically reduce the time spent on segmenting health insurance claims in health insurance administration in Nigeria. The proposed algorithms for the improved K-means Clustering are implemented using JAVA. The improved K-means clustering algorithm was employed in solving the segmenting problems of the health insurance claims and iris dataset. This study concluded that the proposed algorithms are superior to the traditional K-means clustering and the simple K-means in WEKA in terms of convergence and accuracy. The resulting clusters will be used in further studies for developing the supervised classification approach.

K-means clustering, Distance functions, Euclidean Distance, Manhattan Distance, health insurance, health insurance claims.

Haar wavelet is one of the best mathematical tools in image cryptography and analysis. Because of the specific structure, this wavelet has the ability which is combined with other mathematical tools such as chaotic maps. The rational order chaotic maps are one of clusters of chaotic maps which their deterministic behaviors have high sensitivity. In this paper, we propose a novel method of gradient Haar wavelet transform for image encryption. This method use linearity properties of the scaling function of the gradient Haar wavelet and deterministic behaviors of rational order chaotic maps in order to generate encrypted images with high security factor. The security of the encrypted images is evaluated by the key space analysis, the correlation coefficient analysis, and differential attack. The method could be used in other fields such as image and signal processing.

Cryptography; image encryption; Wavelets; Rational order chaotic maps; Chaos.

Computational Grids is highly heterogeneous shared resources for problem solving in any dynamic environment. Accepting Grid computing technologies will be difficult, unless users are certain of safety of their data like in their own environment. Security in Computational Grids is in two folds. Security of the grid users ensuring authentication, confidentiality, integrity, single sign on and delegation on one hand and security of the grid resources in the form of authorization and access control on the other hand. Existing methods of authentication in computational grids have proved inadequate for identifying users, hence, a more reliable technique is required. This paper provides a model, which allows users reliable transactions in grid by using fingerprint to enhance security in grid. Thus, this study aimed at hybridizing fingerprint biometric and Attribute Based Access Control (ABAC) for authenticating and authorizing computational grid users based on attributes of the users for computational grid resources.

Authentication, authorization, grids, biometric, security.

This project is designed to develop Gesture Recognition Robot via Kinect Sensor. The idea of this project is starting from research of Kinect on pc. This paper is to provide plenty useful feature of Kinect such as controlling robot by using our gesture and motion. It is a breakthrough in the market as many applications will be using gesture or motion to control it. This project is a startup for various applications of Kinect or any sensors with the same capability to improve daily life.

Gesture Recognition, Kinect, Robot.

This article presents a state of the art on the main multidimensional indexing methods for structuring the description space still images. We present some general principle, the strengths and weaknesses of these techniques. There are two main families of multidimensional indexing methods: conventional indexing methods and indexing methods based on filtering also called approximation approach. In this article we present only conventional indexing methods.

Data Mining, multidimensional indexing, M-Tree, K-Tree.