Volume 08 Issue 04, July 2020

Optimizing Initial Cluster Center Based on Data Analysis Using K-Means Clustering Algorithm and PAVSM

Chetali Makode, Kedar Nath Singh

Page no:01-06


As a partition primarily based bunch algorithmic rule, K-Means is wide utilized in several areas for the options of its efficiency and simply understood. However, it's documented that the K-Means algorithmic rule could get suboptimal solutions, depending on the selection of the initial cluster centers. During this paper, they propose a projection-based K-Means data format algorithmic rule. The planned algorithmic rule initially uses a typical mathematician kernel density estimation technique to search out the extremely density information areas in one dimension. Then the projection step is to iteratively use density estimation from the lower variance dimensions to the upper variance ones till all the scale is computed. Experiments on actual datasets show that our technique will get similar results compared with different typical strategies with fewer computation tasks this paper reviews numerous strategies and techniques utilized in literature and its benefits and limitations, to research the more would like of improvement of the k-means algorithmic rule. Planned algorithmic rule (PAVSM) increased information analysis.

Protection of System from Data Breaches

Ankur Gupta, Bhupendra Malviya, shital Gupta

Page no:07-10


the tremendous increase in computer users, internet & cyberspace is giving rise to more number to cybercrimes. Technocrats or popularly known as cybercriminals make use of technology, social engineering & other techniques to extract the confidential information. So there is a new need to have a comprehensive understanding of cyber-attacks and its classification & how one can get secured. In this paper, we provided a noble metric concept to test data breaches algorithm. Cybersecurity ensures the protection of information systems including software, hardware and information (data). The purpose of this paper is to give a review on Cybersecurity, its goals, impacts, issues and noble concept. The article also includes a brief description of various types of data breaches that have occurred in the past.

Review of Face Recognition System

Md. Azheruddin, Sudhir Goswami, Shital Gupta

Page no:11-15


Face recognition is one of the most suitable applications of image analysis. It’s a true challenge to build an automated system which equals human ability to recognize faces. While traditional face recognition is typically based on still images, face recognition from video sequences has become popular recently due to more abundant information than still images. Video-based face recognition has been one of the hot topics in the field of pattern recognition in the last few decades. This paper presents an overview of face recognition scenarios and video-based face recognition system architecture and various approaches are used in video-based face recognition system which can not only discover more space-time semantic information hidden in video face sequence, but also make full use of the high level semantic concepts and the intrinsic nonlinear structure information to extract discriminative manifold features. We also compare our algorithm with other algorithms on our own database.

Intrusion Detection System using Particle Swarm Optimization Algorithm

Manjeet Kumar Soni

Page no:16-20


We see security issues in mobile networks from different malicious attacks. These attacks may be of different types, so the security of the system needs different ways to manage them. By this, the system varies in losses like data, energy, and efficiency. Here we presented an intrusion detection system that uses feature optimisation, feature extraction, and classification techniques. Here we calculated the direct and indirect trust values from every node and calculated the trust feature. The Particle Swarm Optimization algorithm (PSO) is used for feature optimisation, and each node of MANET is used to extract the trust features. The obtained optimised features are then classified with the Neural Network (NN) classifier, which detects the intruder. The parameters like successful packet delivery, communication delay, and the required energy consumption for the identification and isolation of intruders are used for evaluation and comparison. The optimisation methodology, which does not use feature optimisation, has achieved the PDR as 89%, 22.45 ms of latency, and 180.7 mJ of energy consumption at 10% of the MANET's malicious nodes. The work which uses feature optimisation has achieved 97% of PDR, 10.15 ms of latency and 128.8 mJ of energy consumption at 10% of malicious nodes present in manet.