International Journal of Advanced Networking and Applications

Hesham Mohammed Ali Abdullah ¹, A. V. Senthil Kumar ²

Affiliations
1 PG and Research Department of Computer Application, Hindusthan College of Arts and Science, Coimbatore, IN
2 PG and Research Department of Computer Applications, Hindusthan College of Arts and Science, Coimbatore, IN

Abstract

Cognitive Radio (CR) technology has significant impacts on upper layer performance in Ad Hoc Networks (AHNs). In recent times, several number of investigation are conducted in CR are mostly focusing on the opportunistic spectrum admission and physical layer transmission throughput. However, CR technology determination also have considerable impacts in mobile Ad Hoc networks (AHNs), which enables wireless devices to dynamically create networks without essentially use of a fixed infrastructure . Nowadays, establishing a cognitive network is such a difficult task. The most important issues is routing in CRAHNs. In this paper, it majorly focuses on the survey of routing and opportunistic routing schema in CRAHN. The most significant scheme behind this concept is to make use of a suitable routing protocol designed for establishing Cognitive Radio Network (CRN). Due to licensing, the accessibility of radio frequency for wireless communication gets reduced day by day. Thus, there is a necessitate to have some other way to use these frequencies in an efficient manner. Routing is efficient method to solve these issues, but the use of geographical concept is also a challenging task in CRN. Since, there is a lack in detailed understanding of these extremely dynamic opportunistic links and a consistent end-to-end transportation mechanism over the network. Here, it focuses on the study of possible routing approaches with the purpose of be able to be employed in CRAHNs. There is a comparison on performance evaluation of various potential routing approaches in terms of table significant reduction and what solution can be found from the routing protocol are also discussed. The routing protocol attains reliable communications for CRAHNs, without usually getting feedback information from nodes in a CRAHN to considerably accumulate the communication overhead.

Keywords

Cognitive Radio (CR), CR Routing Protocol (CRP), CR Ad Hoc Network (CRAHN), Dynamic Spectrum Access (DSA), Spread Spectrum, Opportunistic Routing, classical Routing Schema and Spectrum Sharing.

Full Text

   

K. Shanmuga Priya ¹, A. V. Senthil Kumar ²

Affiliations
1 Department of Computer Science, Hindusthan College of Arts and Science, Coimbatore-28, IN
2 Department of MCA, Hindusthan College, Coimbatore-28, IN

Abstract

The main goal of educational data mining is to improve the student performance. The usage of data mining techniques which achieves the goal in an efficient manner. The discovery of knowledge that extract from the end semester [1] is one of the method for improving the quality of higher education. In the higher education, the analysis on enrolment of student's performance in a particular course, the student talent, confidence, studies and ethic helps to get more knowledge. In this research, the data classification and decision tree [1] which helps to improve the student's performance in a better way. But with the inclusion of extracurricular activities with the above data mining techniques makes quality of education in an easiest way. This type of approach gives high confidence to students in their studies. This method helps to identify the students who need special advising or counseling by the teacher which gives high quality of education.

Keywords

Classification, Educational Data Mining (EDM), ID3 Algorithm, Knowledge Discovery in Database (KDD).

Full Text

   

A. V. Senthil Kumar ¹, S. Mythili ¹

Affiliations
1 Department of Computer Applications, Hindusthan College of Arts and Science, Coimbatore, IN

Abstract

The existing clustering algorithm has a sequential execution of the data. The speed of the execution is very less and more time is taken for the execution of a single data. A new algorithm Parallel Implementation of Genetic Algorithm using KMeans Clustering (PIGAKM) is proposed to overcome the existing algorithm. PIGAKM is inspired by using KM clustering over GA. This process indicates that, while using KM algorithm, it covers the local minima and it initialization is normally done randomly, by KM and GA. It always converge the global optimum eventually by PIGAKM. To speed up GA process, the evalution is done parallely not individually. To show the performance and efficiency of this algorithms, the comparative study of this algorithm has been done.

Keywords

Clustering, Genetic Algorithm, K-Means, Mutation, Parallel.

Full Text

   

B. Kaviyarasu ¹, A. V. Senthil Kumar ¹

Affiliations
1 PG and Research Department of Computer Applications, Hindusthan College of Arts and Science, Coimbatore-38, IN

Abstract

Predicting the objective of internet users has divergent applications in the areas such as e-commerce, entertainment in online, and several internet-based applications. The critical part of the classifying internet queries based on obtainable features namely contextual information, keywords and their semantic relationships. This research paper presents an improved support vector machine classifier that makes use of ad boost genetic algorithmic approach towards web interaction mining. Around 31 participants are chosen and given topics to search web contents. Parameters such as precision, recall and F1 score are taken for comparing the proposed classifier with the classical support vector machine. Results proved that the proposed classifier achieves better performance than that of the conventional SVM.

Keywords

Web Interaction Mining, Algorithm, Support Vector Machine, Classifier, Ad Boost, Genetic Algorithm.

Full Text

   

M. Kalpana ¹, A. V. Senthil Kumar ²

Affiliations
1 Department of Physical Science and Information Technology, Agricultural Engineering College and Research Institute, Tamil Nadu Agricultural University, Coimbatore-641003, IN
2 Department of MCA, Hindusthan College of Arts & Science, Behind Nava India, Coimbatore-641028, IN

Abstract

The fuzzy logic and expert system is an important technique to enhance the machine learning reasoning. In this paper, we propose a fuzzy expert system framework which constructs large scale knowledge based system effectively for diabetes. The knowledge is constructed by using the fuzzification to convert crisp values into fuzzy values. By applying the fuzzy verdict mechanism, diagnosis of diabetes becomes simple for medical practitioners. Fuzzy verdict mechanism uses triangular membership function with mamdani's inference. Defuzzification method is adopted to convert the fuzzy values into crisp values. The result of the proposed method is compared with earlier method using accuracy as metrics. The proposed fuzzy expert system can work more effectively for diabetes application and also improves the accuracy of fuzzy expert system.

Keywords

Diabetes Application, Fuzzification, Fuzzy Verdict Mechanism, Fuzzy Expert System.

Full Text