Santosh Kumar
@viit.ac.in
Assistant Professor, AI&DS
Vishwakarma Institute of Information Technology, Pune
Pursuing PhD Computer Science & Engineering - Data Science at KLEF, Vijayawada, where I learned the intricacies of Machine Learning modelling, Natural language Processing and Statistical Analysis of Health care informatics.
Ease in leading as well as working in teams and sound knowledge on work experience in data analysis, interpret and suggest solutions to correct the deviations by utilizing the statistical analysis concept and deliver a quality result.
Willingness to learn and take on newer challenges is my passion.
EDUCATION
M.Tech(CSE)
RESEARCH INTERESTS
Machine Learning, High performance Computing, Cloud & Devops
Scopus Publications
Scopus Publications
Nirupma Pathak, Bandan Kumar Bhoi, Neeraj Kumar Misra, and Santosh Kumar
World Scientific Pub Co Pte Ltd
As the semiconductor industry strives for downsizing and high speed, it is confronted with increasing scaling uncertainty as devices decrease to the nanoscale. Nano-magnetic logic (NML) is an alternative approach to synthesize the digital logic circuits with high-density and low-power consumption. We introduced an optimal design of content addressable memory (CAM) memory based on perpendicular nano-magnetic logic (pNML). The main aim of this implementation is to synthesize CAM memory in terms of latency and other design parameters. The implementation of the design is a multilayer approach, which is optimal. The synthesis approach and optimization are perfectly scalable across layout construction of designs. Here a new logic gate in pNML technology is designed which is mainly used for matching of two input numbers. According to insight, both memory unit and a matching unit in the pNML are introduced in the state-of-the-artwork for the first time to synthesize design in high-speed pNML application. MAGCAD tool is used for the design of all the proposed pNML layouts.
Nirupma Pathak, Neeraj Kumar Misra, Bandan Kumar Bhoi, and Santosh Kumar
Springer Singapore
Nirupma Pathak, Neeraj Kumar Misra, Bandan Kumar Bhoi, and Santosh Kumar
Allerton Press
Praful Saxena and Santosh Kumar
IEEE
Embedding the Watermark in all frames of cover video is not a good practice in terms of complexity . This paper proposed the method in which the watermark will not be inserted into all frames of cover video. Due to the frequent illegal copying of digital media like image , video and audio. Watermarking method is playing an important approach for the purpose of copy right protection. This paper propose SCD method which enables to embed the watermark in selected frames of cover . Use of wavelet transform technique and singular value decomposition makes the approach robust against various attacks . The performance measurement of the algorithm shows the robustness against intentional and unintentional attacks.
Bandan Kumar Bhoi, Nirupma Pathak, Santosh Kumar, and Neeraj Kumar Misra
Springer Science and Business Media LLC
S.Sathish Kumar and Aruchamy Rajini
IEEE
Nowadays, peoples are posting their personal opinion about different brands on social media. Brands such as airlines, cars, sports teams, entertainment, finance, retail, and the food industry are popular on Twitter. Among them, airlines are the most used brand in business people. In airline companies, building customer loyalty is a major challenge. In the past, traveller's reviews are manually read, analyze, and categorize, which required a lot of effort and time. To avoid the problem, an efficient automatic sentimental analysis method is proposed. The main aim of this paper is to fining best and work airline from tweets. The proposed method consists of three stages namely, pre-processing, feature extraction, and sentiment classification. Initially, the collected reviews are pre-processed to reduce the complexity of classification process. After the pre-processing, word2vec conversion is performed using a skip-gram model. Then this output vector is given to the Adaptive Rider optimization based support vector neural network (AR-SVNN) classifier to classify a review as positive, negative, or neutral. Here, the SVNN classifier is enhanced using adaptive rider optimization (ARO) algorithm. The performance of the proposed methodology is analyzed in terms of different metrics namely, accuracy, precision, recall and F-Measure. The experimental results show that proposed method outperformed the other methods.
Santosh Kumar and T. Vijaya Shardhi
Springer Singapore
Nirupma Pathak, Santosh Kumar, Neeraj Kumar Misra, and Bandan Kumar Bhoi
Springer Science and Business Media LLC
Abstract Quantum technology has an attractive application nowadays for its minimizing the energy dissipation, which is a prominent part of any system-level design. In this article, the significant module of a multiplexer, an extended to n:1 is framed with prominent application in the control unit of the processor. The proposed multiplexer modules are framed by the algorithm, which is extended perspective based. Further, quantum cost and gate count are less to ensure the efficient quantum computing framed. In addition, the QCA computing framework is an attempt to synthesize the optimal primitives in conservative reversible multiplexer in nano-electronic confine application. The developed lemmas is framed to prove the optimal parameters in the reversible circuit. Compared with existing state-of-art-works, the proposed modular multiplexer, the gate count, quantum cost and unit delay are optimal.
With an enhancement in technology and development of sensors, they have decided to concern the latest technology in particular areas to increases the Quality of their life. One of the important section of research that has seen an agreement of the latest technology is the healthcare. the people who need healthcare services they have to suffer for long process, to reduce the problem As a result, basically this complete technology would try to resole the healthcare issues. The main goal of the project is to develop a remote healthcare system. It have three main parts. The first and most important section in this process is to detecting the patient current health status using this sensors. second important thing is sending data to cloud storage and the last section is to get the detected data for remote viewing. Remote viewing of the data provided to the doctor will able to monitor a patients health progress aboard from hospital places. The Edge-Cognitive- Computing-based (ECC-based) smart healthcare system. This technology is able to check also distinguish and detect the current status of patient using this sensors. It also accommodate the computing resource appropriation of the complete edge computing network expansive to the health different percentage of each user. The experiments show that the ECC-based healthcare system provide a exceed user background and try to improvement the computing resources moderately, as well as consequentially improving in the durability rates of patients in emergency
S.Sathish Kumar and Aruchamy Rajini
IEEE
Data Mining is a process of generating new information from the existing datasets involving the machine learning, statistics and database systems. Major role of data mining is to turn raw data into useful information by using mathematical algorithms for predicting future events. Data mining is known as knowledge discovery in data (KDD). The key properties of data mining are automatic discovery of patterns, prediction of likely outcomes, creation of actionable information and focuses on large datasets and database. Mining the information related to a content-oriented extraction on the basics of feature of data. For accessing the features of data that is data retrieval can be done by feature extraction mechanism. Various feature extraction methods have been used along with a feature selection algorithm for mining the large datasets.
S. Sathish Kumar, P. Manimegalai, and S. Karthik
IEEE
Compact exceptionally delegated frameworks (MANETs) are without establishment frameworks prompted by remote adaptable contraptions with kept battery life. This enveloped battery confines in MANETs makes it required to consider the imperativeness mindfulness feature in their layout. In perspective of the way that guiding traditions have central part in MANETs, their essentialness mindfulness grows orchestrate life time by capably utilizing of the open limited imperativeness. TORA is one of these directing traditions that offer abnormal state of flexibility. This paper uses the EBC (Efficient Binary Calculation) to organize the vitality mindfulness feature to the TORA directing tradition. The proposed tradition considers courses length in its course isolate plans and furthermore fuses courses imperativeness level in its figures. It figures the directing issue as a change predicament and thereafter uses EBC to sedative a course that extends a weighted limit of the course length and the course imperativeness level. Expansive amusements in ns-2 test framework condition exhibit that the proposed directing tradition, called EBC-TORA, expands the framework lifetime shockingly and beats TORA to the extent framework life time, structure life time and total appropriated data.
Vandana B. and S. Sathish Kumar
Author(s)
Agriculture is the primary concern in India and it requires more technological developments to be incorporated when compared to other industrial developments. Information and Communication Technologies (ICT) provides a simple and cost effective technique to the farmers for enabling the precision agriculture. An information sharing system is developed through this study to support the small scale farmers. The technological developments in smart phone application and web application are used as a path to reach the small scale farmers. The application developed through this study shares an information about sugar cane crop like fertilizer requirement, pesticide requirement etc. This system will provide important and timely information to the small scale farmers like a handy agriculture guide.Agriculture is the primary concern in India and it requires more technological developments to be incorporated when compared to other industrial developments. Information and Communication Technologies (ICT) provides a simple and cost effective technique to the farmers for enabling the precision agriculture. An information sharing system is developed through this study to support the small scale farmers. The technological developments in smart phone application and web application are used as a path to reach the small scale farmers. The application developed through this study shares an information about sugar cane crop like fertilizer requirement, pesticide requirement etc. This system will provide important and timely information to the small scale farmers like a handy agriculture guide.
B Vandana and S Sathish Kumar
IEEE
Agriculture is the main work field in India. Farming industry adopts less innovative technology compared to other industries. Information and Communication Technologies provides simple and cost effective techniques for farmers to enable precision agriculture. The work propose a state of the art model in agriculture field which will guide the rural farmers to use Information and Communication technologies (ICT) in agriculture fields. Big data analytics is used to improve the crop yield. It can be customized for precision agriculture to improve the quality of crops which improves the overall production rate.
Santosh Kumar, Atul Chaudhary, Ravindra Singh, Manish Sharma, and Anil Dubey
Springer Singapore
T. V. Radhika, Krushna Chandra Gouda, and S. Sathish Kumar
IEEE
Currently climate research is most priority area as climate change generally affects the society as a whole. So there is requirement to study the weather and climate variability at very high resolution in multiple spatial and temporal scales. Presently climate data are huge in size as more data are being generated compared to past. Also sophisticated models are being used for prediction of weather and climate variability, generating vast amount of multidimensional digital data. As such climate data is considered to be big data which is multi-dimensional, multi-approach and multisource. So the demand of high performance computing and cloud computing is being increased to do the climate research. This paper gives an outline of a few strategies in supporting big data administration and investigation in geoscience domain for climate studies. By analyzing contemporary information technologies and approaches, it can confirm what operational program framework and approaches are at hand and pertinent in developing big data-driven climate research. A transitory overview of HBase for storing and managing notable geoscience data across distributed machinery is highlighted. Withal MapReduce-predicated techniques to fortify parallel access of massive NetCDF data are considered. The outcomes can recognize basic issues and enhances the proficiency of dissecting huge geoscience information by lessening information preparing time.
Ujwal Arjun Bodke and Santosh Kumar
IEEE
XML actually developed as a benchmark for caching, dispense and interchanging data over multiple platforms. The XML data is on the grow over the time in fast rate. Enterprises want formulating queries on XML datasets habitually. As giant XML data is retrievable, it is not easy job to pull out vital data from XML dataset. It is computationally expensive to answer queries without any sustain. Towards this, in this paper we introduce a method known as Tree-based Association Rules (TARs) mined policy that make available necessary info on structure and content of XML file and the TARs are also placed in XML design. The extracted data (TARs) used in future on for XML query answering support system. This technique gives rapid and precise answer. We have also introduced a model to show the usefulness of the designed structure. The experimental outcomes are very accommodating and query answering is likely to be important in real time uses.
Prajakta R. Bhamre, Rutuja G. Chauhan, and Santosh Kumar
IEEE
In this paper the Laser Gesture Detection Technique is proposed. In today's modern world it is a need of time to complete each and every work quick and fast. Proposed system is used to perform day to day activity faster and with fewer labors to make human life luxuries and physically disabled human to perform their task easily. This technique is based on laser image capturing performed by operator on screen by laser device with high resolution camera followed by image processing, color filtration and depth histogram with the help of algorithm for space discretization for pattern normalization. Normalized pattern will be again filtered for binary matching against gesture template database. Really Confirmation will be performing the final task after successfully binary matching of images with database.
Manish Sharma, Atul Chaudhary, Manish Mathuria, Shalini Chaudhary, and Santosh Kumar
IEEE
In many organizations large amount of data are collected. These data are sometimes used by the organizations for data mining tasks. However, the data collected may contain private or sensitive information which should be protected. Privacy protection is an important issue if we release data for the mining or sharing purpose. Privacy preserving data mining techniques allow publishing data for the mining purpose while at the same time preserve the private information of the individuals. Many techniques have been proposed for privacy preservation but they suffer from various types of attacks and information loss. In this paper we proposed an efficient approach for privacy preservation in data mining. Our technique protects the sensitive data with less information loss which increase data usability and also prevent the sensitive data for various types of attack. Data can also be reconstructed using our proposed technique.