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

Medical imaging systems contribute much towards effective decision-making by the physicians, which is highly essential in the day-to-day life of humans. In this study, Taylor-Krill herd (KH)-based support vector machine (SVM) is proposed for medical image denoising. The Taylor-KH-based SVM is the integration of Taylor series in KH optimisation algorithm, which is used for tuning the optimal weights of the SVM classifier. The efficiency of KH is due to two global and two local optimisers, and the adaptive operators ensure the adaptive nature of KH. Above all, KH never uses the derivative information as it employs the stochastic search and thereby, reduces the complexity of the algorithm. The proposed method tunes the hyperplane parameters of SVM optimally so that the optimal identification of the noisy pixels in the image is ensured and replaced with adaptive weights. The proposed method is analysed based on the metrics, such as peak signal-to-noise ratio (PSNR), structural similarity (SSIM) and the comparative analysis is done with existing methods for showing the effectiveness of the proposed method. The simulation result shows that the proposed method acquired a PSNR of 30.36 dB and SSIM of 0.89, respectively. © The Institution of Engineering and Technology 2019

World University root node

Department of Computational Intelligence

School of Computer Science and Engineering

Vellore Campus

IET Image Processing

Integrating Taylor-Krill herd-based SVM to fuzzy-based adaptive filter for medical image denoising

The compatibility issues among the characteristics of data involving numerical as well as categorical attributes (mixed) laid many challenges in pattern recognition field. Clustering is often used to group identical elements and to find structures out of data. However, clustering categorical data poses some notable challenges. Particularly clustering diversified (mixed) data constitute bigger challenges because of its range of attributes. Computations on such data are merely too complex to match the scales of numerical and categorical values due to its ranges and conversions. This chapter is intended to cover literature clustering algorithms in the context of mixed attribute unlabelled data. Further, this chapter will cover the types and state of the art methodologies that help in separating data by satisfying inter and intracluster similarity. This chapter further identifies challenges and Future research directions of state-of-the-art clustering algorithms with notable research gaps.

VIT Unitversity

Pattern Recognition Applications in Engineering Advances in Computer and Electrical Engineering

A Study on Efficient Clustering Techniques Involved in Dealing With Diverse Attribute Data

Brain tumor diagnosis using MRI is an active area of research and the traditional methods of brain tumor classification lack the accurate identification of the tumor region, which demands an effective classification strategy for brain tumor identification. Accordingly, an effective classifier, named as Chaotic Biogeography-based Ride-Neural Network (CBB Ride-NN) is proposed for brain tumor detection using MRI. Initially, the brain tumor segmentation is processed using the active contour model, where the tumor and the non-tumor regions are differentiated from the brain images in order to extract the texture features for classification. Thus, for classification of brain tumors, proposed Chaotic Biogeography-based Ride-Neural Network is employed, where the proposed CBB Ride optimization algorithm trains the neural network. The classification accuracy of the brain tumor classification is effective and addresses the issues of the existing classifiers. The effectiveness of the proposed CBB Ride-NN for tumor classification is revealed using the analysis with the BRATS dataset with respect to the performance metrics, namely accuracy, specificity, and sensitivity. The measures, like accuracy, sensitivity, and specificity acquired the maximal values of 0.9327, 0.9288, and 0.9226, respectively. © 2020, Institute of Advanced Scientific Research, Inc.. All rights reserved.

Journal of Advanced Research in Dynamical and Control Systems

Chaotic biogeography rider-based neural network for brain tumor classification using MRI images

The cell adequacy is one of the important factor is Pap smear test as per Bethesda reporting system. The adequate number of cells in a given sample helps to effectively check for abnormality of cells. The Pap smear sample with inadequate number of cells will be rejected and patient will be called for collecting sample once again. The cell adequacy is the number of squamous epithelial cells present in the given Pap smear sample. This paper presents an automated method which estimates the number of cells in a given Pap smear image. This method estimates a cell count even when cells are overlapped. The proposed approach first separates disjoint cells and overlapped cell cluster from cervical Pap smear image. Then the overlapped cluster is further segmented to obtain individual cells. The overlapped cluster is selected based on the area factor. That is area of overlapped cluster will be more than disjoint cells. This method uses Bitplane slicing technique for segmentation of overlapped cells. Then the total cell counts is estimated as sum of disjoint cells count and number of cells separated from overlapped cluster. The method is tested on the ThinPrep images publicly available in open access database.

International Journal of Recent Technology and Engineering Regular Issue

Estimating Cell Count in Overlapped Cluster of Thin Prep Pap Smear Cell Microscopic Image

The recommender system plays a vital role in the field of online business. To improve the quality of the recommender system, in this research work we focus on enhancingthesentiment analysis process to generate the most efficient recommendations to the users. E-commerce website has given consumers a venue to share their opinion on different entities. These opinions have a high impact on other customers to decide whether to purchase a product or not. By taking advantage of this, some users follow unethical ways to promote their products. Analyzing these opinions manually is difficult, and thus Attention-based Bi-LSTM ConvolutionNeural Network (ABC) with RMSprop optimizer is proposed to extract the genuine opinion level of reviews automatically. The contributions are: (i) Development of the ABC model for text classification with fake review detection, (ii) Evaluation of the results in cross-domain datasets, (iii)Analyzing the impact of different optimization algorithms in the proposed neural networkthat can improve the classification task, (iv) Finally, sentiment analysis module is passed to recommender module to produce Top N recommendations.Extensive experiments are carried out on three different domain datasets and the results clearly indicate ABCwith RMSprop optimizer outperforms better than state-of-art text classification methods. © 2019, Institute of Advanced Scientific Research, Inc.. All rights reserved.

An improved personalized recommender system with optimized ABC neural network

Predicting the most probable word for immediate selection is one of the most valuable technique for enhancing the communication experience. With growth in mobile technologies and vast spread of the internet, socializing has become much easier. People around the world spend more and more time on their mobile devices for email, social networking, banking and a variety of other activities. Due to fast paced nature of such conversation it is necessary to save as much as time possible while typing. Hence a predictive text application is necessary for this. Text prediction is one of the most commonly used techniques for increasing the rate of communication. However, the speed at which text is predicted is also very important in this case. The objective of this work is to design and implement a new word predictor algorithm that suggests words that are grammatically more appropriate, with a lower load for system and significantly reduce the amount of keystrokes required by users. The predictor uses a probabilistic language model based on the methodology of the N-Grams for text prediction. © BEIESP.

International Journal of Innovative Technology and Exploring Engineering

Real time word prediction using N-grams model

Since last decade, due to increasing demand, huge amount of software is being developed, whereas the data intensive applications have also increased the complexity in these types of systems. Also, during the development process, software bugs may severely impact the growth of industries. Hence, the development of bug free software application is highly recommended in the real-time systems. Several approaches have been developed recently that are based on the manual inspection but those techniques are not recommended for huge software development scenario due to maximum chances of error during manual inspection. Thus, machine learning based data mining techniques has gained huge attraction from researchers due to their analyzing and efficiently detect the defect by learning the different attributes. In this work, we present machine learning based approach for software defect prediction. However, software defect datasets suffer from the high dimensionality issues, thus we present a novel subspace clustering approach using evolutionary computation based optimal solution identification for dimension reduction. Later, Support Vector Machine Classification scheme is implemented to obtain the defect prediction performance. Proposed approach is implemented using MATLAB simulation tool by considering NASA software defect dataset. A comparative study is presented which shows that proposed approach achieves better performance when compared with the existing techniques. © 2005 – ongoing JATIT & LLS.

Journal of Theoretical and Applied Information Technology

An evolutionary approach for software defect prediction on high dimensional data using subspace clustering and machine learning

The power of Deep Learning Networks has allowed us to build applications far beyond what was thought possible during our time. But the basic forms of these architectures still have their limitations in terms of the data needed and difficulty in understanding and tuning the parameters of these networks. Our project aims to deal with these limitations as effectively as possible. For this reason we have chosen to implement a Siamese Neural Network in order to overcome the requirements of classical Deep Learning based image classification. And in order to further increase the efficiency of the network, we will process it using a genetic algorithm and harness it to improve the architecture and training efficiency by letting the algorithm fine-tune the parameters to get the best possible configuration for the neural network. © BEIESP.

International Journal of Engineering and Advanced Technology

Optimization of siamese neural networks using genetic algorithm

Implementation of advanced analytics for big data processing in business intelligence is significant towards gaining profits. For processing large-scale data sets efficiently, so many challenges are faced by traditional database system. To overcome the disadvantages present in an existing system, various kinds of a new database have been evolved along with application programs (e.g., MySQL, PostgreSQL, Hive, etc.). Such type of systems store the data in the database, retrieves, and displays the information once it is queried. The time duration varies in the different database for doing the process. This paper evaluates the performance by using SQL*Plus and Hive. In the enterprise business data model, a comparison of both SQL*Plus and Hive for some CRUD operations (Insert, Join, and Retrieve) are estimated. By the work presented in this paper, we conclude if performance is a key, then SQL*Plus is a right choice to use and for processing large datasets hive works better. © 2019, Springer Nature Singapore Pte Ltd.

Department of Software Systems

Advances in Intelligent Systems and Computing Advances in Big Data and Cloud Computing

Evaluating the Performance of SQL*Plus with Hive for Business

2018 4th International Conference on Computing Communication and Automation (ICCCA)

Weighted Finite Automata for Regularity Recognition in Textures

Programming measurements was utilized for foreseeing issue in modules of programming ventures. Convenient forecast of flaws enhances programming quality and subsequently its dependability. In this paper, a framework towards subspace grouping of large data set was proposed at class level to minimize the error. We composed an iterative calculation for grouping of high dimensional datasets for improvement of a target work. At that point the bunched data sets were examined utilizing Step-Wise Linear Regression to investigate the relationship among a structure variable and the autonomous factors in order to anticipate of damaged and non-faulty classes. To evaluate the supportiveness of the model, we drove a practical learning on the Attitude Survey Data. The proposed strategy specifically managed blunder variables and consequently gave precise fault prediction least standard error (0.003) when contrasted with the current technique (4.687). Root mean square error which measures the distinction between the assessed error and the real error was (0.8) in the proposed technique. The results demonstrated that the forecast models based on subspace clustering were essentially predominant to the current techniques. © 2018 Sumangala Patil et. al.

International Journal of Engineering and Technology(UAE)

Class level software fault prediction using step wise linear regression

Cells mostly keep on changing on various kinds of medical factors. At initial stages, the identification of any abnormality or cancer in cell is more important. The increasing population somehow raised awareness for maintaining better health. Because of degrading quality of food, smoking habits and most importantly pollution of environment are the cause of cell damages or abnormalities of tissues among the people. So, for providing better medical treatment for abnormal cell diseases, it is better to identify the affected cell at its initial stage. For efficient analysis, it is also important to detect the affected cells at an initial stage. Thus, the patients will get the best solution for the diseases. Previously, researchers used various methods as for enhancing images such as Fast Fourier Transform method, Binarization for extracting images etc. Here, we are going to use watershed processing for segmentation which can easily detect abnormalities in cell quickly. © RJPT All right reserved.

Research Journal of Pharmacy and Technology

Abnormal Lung Cells Detection using Watershed Algorithm

Image enhancement plays a key role in mammography for better analysis of breast cancer detection. Poor quality of mammogram images may produce ambiguous assumptions about the results in breast cancer predictions. Existing image enhancement methods such as median filter, morphological method, wavelet transformation, contrast stretching are produced better quality of images, however, they cannot accentuates the image features includes boundaries and edges. In mammography, it is required to accentuate the boundary and edges features with good quality. Therefore, this paper proposed the work for intelligibility of these boundary and edges features, known as intelligibility mammogram enhancement method (IMEM). Demonstration of efficiency of proposed method and comparative results are presented in the experimental study using mammogram image datasets. © 2018 IEEE.

2018 2nd International Conference on Inventive Systems and Control (ICISC)

Improving the mammogram images by intelligibility mammogram enhancement method

Before chatbots there were simply bots: The invention of a chatbot brought us to the new era of technology, the era of conversation service. A chatbot is a virtual person that can effectively talk to any human being with the help of interactive conversion textual skill. Now a days there are many cloud-based platforms available for developing and deploying the chatbot such as Microsoft bot framework, IBM Watson, Kore, AWS lambda, Microsoft Azure bot service, Chatfuel, Heroku and many more but all those techniques has some drawbacks such as built-in Artificial Intelligence, NLP, conversion service, programming etc. This paper represents the comparison between all cloud-based chatbot technologies with some constraint such as built-in AI, setup time, completion time, complexity etc. Finally, by the comparison, we will get to know that which cloud platform is efficient and suitable for developing chatbot.

Department of Instrumentation

School of Electrical Engineering

Fulltext

International Journal of Engineering & Technology

Comparative study of cloud platforms to develop a Chatbot

Development of software passes through different stage of evolution which is called as software development cycle. Designing is a critical area in the software development cycle. Designing occurs early in software development life cycle and the errors introduced during the design of the software can make the software totally unusable in production. The design errors must be identified and corrected, so that cost of correcting defects in the later stages of software development cycle is minimum. In this survey paper, we study the current methods to identify software design errors and find the current gaps to propose open areas for further research. © Research India Publications.