Vellore Institute of Technology (VIT) is a private university located in&nbsp;Tamil Nadu, India. Founded in 1984, as Vellore Engineering College, the institution offers 20 undergraduate, 34 postgraduate, four integrated and four research programs. It has campuses in Vellore, Amravati, Bhopal and Chennai.

VIT is one of the top ranked private universities in India according to NIRF, THE and QS Rankings.&nbsp;Govt. of India has recognized&nbsp;VIT, Vellore as an&nbsp;Institution of Eminence. This has allowed VIT to take independent quality initiatives and move up in world ranking.

&nbsp;

&nbsp;

VIT University

In this paper, techniques to improve the overall efficiency of an image encoding algorithm that is specifically designed to be used in Wireless Sensor Networks (WSN) are proposed. The inherent nature of WSN, which contains numerous network nodes each having its computational resources, makes the algorithm is being underutilized due to its sequential nature and hence there is a scope to improve speedup by applying the concepts of parallelization on it. The other avenues which are pertinent to the task are the usage of GPU for computation of large arrays as well as the concept of cluster computing to run the algorithm on a cluster of nodes. Another point of importance for such an algorithm is the utilization of memory and how to make its usage more efficient. Hence, the concept of pre-allocation is also applied to the algorithm to make it run faster. An in-depth study of the algorithm is conducted by considering various parameters such as the size and complexity of the key, histogram analysis, correlation values, etc. © 2021 IEEE.

World University root node

Department of Computational Intelligence

School of Computer Science and Engineering

Vellore Campus

2021 6th International Conference for Convergence in Technology, I2CT 2021

Parallelizing Image Encryption Algorithm for Wireless Sensor Networks

Deep learning comes under a class of machine learning where we use it for extremely high-level output, like recognition of images, etc. It has been used in pattern recognition over a vast area such as handmade crafts to extract the data from learning procedures. At present, it has gained a great significance in robot vision. In this paper, we show how neural networks play a vital role in robot vision. Image segmentation, which is the initial step, is used to preprocess the images and videos. The multilayered artificial neural networks have a lot more applications. It can be applied in drug detection, military bases, and many more. The main objective of this paper is to review how deep learning algorithms and deep nets can be used in various areas of robot vision. There are some predefined deep learning algorithms that are available in the market, which are used here to perform this comparative study. These will help us to have a clear insight while building vision systems using deep learning. © 2021, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

Smart Innovation, Systems and Technologies

Deep Learning for Robot Vision

Document clustering is the most needed process in the data mining field where the number of documents with different methodologies are scattered. The meaningful information can be extracted from the group of documents by grouping them effectively. There are various researches that exist previously which concentrate on clustering the documents present in the real. In the previous works, document clustering is done by using the methodologies called the term weight-based hybridised harmony K-means search (TW HHKM), coverage factor-based hybridised harmony K-means search (CF HHKM), concept-based, kernel and weighted feature-based clustering algorithm (CKW HHKM). Clustering is normally done by using the K-means algorithm and the centroids of clusters are found optimally by using the harmony search algorithm. The problem reside in the above said existing methods are the poor accuracy while clustering the documents where the unrelated documents are grouped together. To overcome this problem, multi-view point HHKM (MP HHKM) approach is introduced, in which clustering can be done accurately. In this work, multi-point analysis is done based on the similarity measurement. The exploratory tests were directed on news group and TREC dataset from which it is robust that the proposed technique MP HHKM overtakes the existing technique with better accuracy values. Copyright © 2021 Inderscience Enterprises Ltd.

International Journal of Cloud Computing

An efficient document clustering using hybridised harmony search K-means algorithm with multi-view point

Nowadays people look for both comfort and secure life whether it is in the house or the workplace as well. The automation system is an advancement in today's technology which is used for controlling home appliances from a far distance to reduce efforts. Security mechanism which has become an important part in today's society is generally used to protect the place. The demand is increasing day by day as it is capable of reducing human work, save daily electricity and also helps to reduce worries about security for families while at work. The focus of this project is to give insights into both automation and security system, which allows us to control door OPEN/LOCK status, gas leakage, and temperature status. There have been many technologies used before and mobile plays a crucial role to control the equipment used in both automation and security system. Through this basic idea, we are trying to bring a safe and monitored view of automation and security systems by making sure that the user is always in control of the house or a vehicle whether it is a city, village or in remote areas. © IAEME Publication.

International Journal of Electrical Engineering and Technology

Automated security systems

The main objective of the proposed methodology is to classify an ECG signal as normal or abnormal using the optimal neuro-fuzzy classifier. The proposed work consists of two phases namely, feature extraction and neuro-fuzzy classifier based classification. The beat signals are initially taken from the physio-bank ATM. Then, three types of features are extracted from each signal namely, Morphological-based features, Haar wavelet-based features, and Tri-spectrum based features. After feature extraction, the optimal neuro-fuzzy classifier is classifying the beat signal as normal or abnormal. Here, Artificial Bee Colony (ABC) algorithm is combined with Genetic Algorithm (GA) for training the neuro-fuzzy classifier. For experimental evaluation, the MIT-BIH Arrhythmia Database is utilized and the performances are analyzed in terms of accuracy, sensitivity, and specificity. The experimental results clearly demonstrated that the proposed technique outperformed by having better accuracy of 93% when compared existing technique achieved 84% only. © 2019, Springer Science+Business Media, LLC, part of Springer Nature.

Multimedia Tools and Applications

GABC based neuro-fuzzy classifier with hybrid features for ECG Beat classification

A recommender system is an information model that ideally, serves context-aware personalized suggestions and recommendation for an entity that the user might want to use or purchase (e.g., music, movies, and books). This addresses the need to filter and prioritize data into germane information for the user. Traditionally, recommender models follow the two most widely used techniques of Content-Based recommendations and Collaborative filtering. A recommender system should be intuitive and understand contextual references. The model should also exhibit serendipity. Which are too accurate fail in practice since our aim is to generate inclusive and non-intrusive recommendations. This paper presents a roundup of the various recommender models and their prediction techniques, aimed at serving as a basic guideline to the people new to the realm of recommender system and to hopefully empower further research and practice in this field. © 2019 American Scientific Publishers

Department of Analytics

VIT Unitversity

Journal of Computational and Theoretical Nanoscience

Qualitative Analysis of Models and Issues in Recommender Systems

Supplementary factor to the general pagerank calculation which is utilized by Google chrome to rank sites in their web index results is tended to in this paper. These extra factors incorporate couple of ideas which expressly results to build the precision of evaluating the PageRank value. By making a decision about the likeness between the web page content with the text extracted from different site pagesresulted in topmost search using few keywords of the considered page for which the rank is to be determined by utilizing a comparability measure. It results with a worth or rate which speaks to the significance or similarity factor. Further, in a similar strategy if sentimental analysis is applied the search results of the keywords could be analysed with keywords of the page considered, it results with a Sentimental Analysed factor.In this way, one can improve and execute the Page ranking procedure which results with a superior accuracy.Hadoop Distributed File System is used to compute the page rank of input nodes. Python is chosen for parallel page rank algorithm that is executed on Hadoop

International Journal of Engineering and Advanced Technology Special Issue

Sentimental Analysis and LSI Similarity Measure for Efficient Page Ranking

The quick development of the Internet surpasses all desires. The investigation and mining of immense sums of web information is confronting a bottleneck in computational power and storage room. Using cloud computing innovation, we can encourage the system access to capable thinking power, stockpiling limit and foundation. Cloud computing can viably take care of the issues by giving an information handling stockpiling focus of high dependability and versatility, which will enhance the capacity to process web information and decrease the necessities of the terminal gadgets. This paper thinks about web mining calculations in a cloud computing platform. The web information mining calculation and the MapReduce programming model are consolidated. We concentrate on the web mining methods, particularly the W-Miner clustering calculation, investigate the mix of web mining calculations and cloud computing innovation and enhance the information mining calculations to adjust to the examination and preparing of mass web information based on cloud computing stages. Our review develops a circulated cloud condition utilizing a Hadoop structure. In the exploratory condition, we dissect the effect on computational execution by setting distinctive piece measure parameters of Prediction accuracy with the number of clusters. Here, the cluster size decides the number that the pending information record is part, and the comparing scale with the measure of Hierarchical Clusters. © 2019 American Scientific Publishers

A Novel Approach for Web Usage Prediction and Recommendation Using Web Session Clustering

One of the curses to the society is a crime which has a deep impact on the society. Victims of crimes are the one who is impacted the most. All communities in the world are affected by crime and the criminal justice system, but largely impacted communities are the backward classes. There are many cases reported of crime committed against scheduled tribes from the year 2005 till date. This paper states the analysis of Crimes Committed against Scheduled Tribes in the year 2015 in various states and union territories in India. In this study, Multiple Linear regression techniques have been used to analyze the crimes committed against scheduled tribes' community in India. This study compares the number of cases reported to the police station and rate of crime committed in different states in India. It also states the future prediction of the crime that would happen. It will also predict the number of cases of crime committed against the scheduled tribe that can be reported in future. The dataset which has been used in this study is taken from official Indian government repository for crimes which include different information of crimes committed against scheduled tribes in different states and union territories measured under the population census of the year 2011. This study will help different Indian states and union territory government to analyze and predict the future crimes that may occur and take appropriate measures against it before the actual crime would occur. © Published under licence by IOP Publishing Ltd.

Department of Software Systems

Fulltext

IOP Conference Series: Materials Science and Engineering

Analysis of crimes committed against scheduled tribes

Clinical decision making is a complex task for physicians as it requires utmost accuracy of diagnosis. This paper proposes a medical decision support system based on least square support vector machine (LS-SVM) and simulated annealing (SA) heuristic for the diagnosis of diabetes in the Pima Indian diabetes (PID) dataset of the UCI machine learning repository. Fisher score (FS) algorithm is used to select the most significant features from the given feature set. LS-SVM with radial basis function (RBF) is used for classification and the SA for optimisation of the kernel parameters of the LS-SVM. The performance of the proposed system is analysed in terms of classification accuracy, sensitivity and specificity using ten-fold cross-validation and confusion matrix. The results show that the classification accuracy of the proposed system outperforms that of various existing systems. © 2016 Inderscience Enterprises Ltd.

International Journal of Information and Communication Technology

An expert system based on LS-SVM and simulated annealing for the diagnosis of diabetes disease

As diagnosing diabetes is a challenging and tedious task for physicians, it has attracted researchers to focus on designing medical decision support systems with higher accuracy. In this paper, a clinical decision support system based on Extreme Learning Machine (ELM) and Simulated Annealing (SA) is proposed for the diagnosis of diabetes using the Pima Indian Diabetes (PID) dataset of the UCI machine learning repository. Classification is performed using ELM while optimization of ELM parameter is carried out by Simulated Annealing heuristics. The performance of the proposed system is analyzed based on the parameters such as classification accuracy, sensitivity and specificity using 10-fold cross-validation and confusion matrix. The accuracy of the proposed system is found to be superior to other existing systems in the literature. The system is validated for different datasets like Hepatitis, Breast Cancer and Cardiac Arrhythmia. © Research India Publications.