MIT-World Peace University (MIT-WPU) is a top educational institute located in Pune, India. It is recognized within&nbsp;UGC under Government of Maharashtra Act No. XXXV 2017.&nbsp;

It is part of the MIT group of institutions (established in 1983) which covers 10+ campuses across India and has more than 50,000+ students enrolled across various disciplines at a given point of time. The institute houses 15 schools ranging from Engineering &amp; Technology, Design, Pharmacy, Commerce, Governance among others. It offers undergraduate, postgraduate and PhD programs among different disciplines.&nbsp;

&nbsp;

MIT World Peace University

School of Computer Engineering & Technology

World University root node

Engineering and Technology

2022 IEEE 7th International conference for Convergence in Technology, I2CT 2022

Real-Time Sign Language Recognition using Deep Learning Techniques

School of Polytechnic and Skill Development

Cybernetics and Systems

Deep E-Learning RecommendNet: An Acute E-Learning Recommendation System with Meta-Heuristic-Based Hybrid Deep Learning Architecture

Individual feature selection algorithms, used for processing high-dimensional multi-source heterogeneous data may lead to weak predictions. The traditional single method process may not ensure the selection of relevant features. The selections of features are susceptible to the changes in input data, and thus fail to perform consistently. These challenges can be overcome by having a robust feature selection algorithm that generates a subset of original features and evaluates the candidate set to check for its relevance. Also, it determines the feasibility of the selected subset of features. The fundamental tasks of selecting feature subset minimize the complexity of the model and help to facilitate the further processing of the model. The limitations of using single feature selection technique can be reduced by combining multiple techniques to generate the effective features. There is a need to design efficient approaches and technique for estimating the feature relevance. This ensemble approach will help to include diversity at input data level, as well as the computational technique. The proposed method---Ensemble Bootstrap Genetic Algorithm (EnBGA)---generates the effective feature subset for the multi-source heterogeneous data. Various univariate and multivariate base selectors are combined together to ensure the robustness and stability of the algorithm. In this pandemic of COVID-19, it's observed that patients already diagnosed with diseases such as diabetes had an increased mortality rate. The proposed method performs feature analysis for such data, where the Genetic Algorithm searches the feature subset and extracts the most relevant features.

Effective Feature Selection Using Ensemble Techniques and Genetic Algorithm

Lecture Notes in Electrical Engineering

Predictive Analytics for Real-Time Agriculture Supply Chain Management: A Novel Pilot Study

2022 IEEE Pune Section International Conference, PuneCon 2022

Personality Prediction with Natural Language Processing using Questionnaire Responses

With constant research in the last decade, Object Detection has become one of the rapidly evolving sub-fields of Deep Learning. As a result, the complexity and applications of Object Detection models have grown, necessitating larger datasets and multi-format labelled annotations for training and testing. To aid with the annotation work, a variety of tools based on various technologies, techniques, and features have been developed. However, some of the tools only cater to specific standards, requirements and problem domain. Researchers have observed that a drop in the quality of labelled annotations can have unintended consequences for model performance. This study provides an overview of various annotation tools with the goal of assisting researchers in determining the appropriate tool for their needs. The work will also serve as a reference for developers as they work on subsequent features to improve or create new tools. © 2022 IEEE.

Proceedings - International Conference on Applied Artificial Intelligence and Computing, ICAAIC 2022

A Review of Image Annotation Tools for Object Detection

In the field of remote sensing satellite imagery, malfunctions in the available raw data are prominent. Especially in Short-Wave Infrared (SWIR) detectors used in satellite imaging cameras, which suffer from dropouts in pixel and line direction in raw data. With the recent development in generative adversarial networks and its vast application in inpainting the missing data, the possibility to predict and fill in the missing data accurately with contextual attention has become prevalent. This paper presents SN-GANs (SN-generative adversarial networks) which is two-staged architecture, and it is based on the concept of feed forward neural networks with contextual attention layers. While reconstructing the corrupted part of the images, the model takes surrounding pixels into consideration. Moreover, this architecture is adept enough to fill in the multiple lines and pixel dropouts efficiently even in super-resolution satellite images. The available traditional methods fail to address the loss of data that incurs, while inpainting a 16-bit raw image because they are effective enough for 8-bit RGB images. SN-GANs have effectively resolved this issue with a lossless image inpainting method for 16-bit satellite images as it retains the features of non-corrupted data. The performance of the model is evaluated using similarity metrics like structural similarity index measure (SSIM), peak signal-to-noise ratio (PSNR) and mean-squared error (MSE). © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

Lecture Notes in Networks and Systems

Reconstruction of Missing Data in Satellite Imagery Using SN-GANs

A huge population of India is dependent on agriculture as their primary source of livelihood. The financial growth of India depends on the GDP (gross domestic product) rate of the agricultural commodities. Since dragon fruit is in high demand both domestically and globally, dragon fruit plantations are economical for small scale to large scale growers and businesses. There is a heavy focus on each stage of the fruit’s development because prolonged or premature harvesting of the fruit can result in food decay and can also impact the India’s economic growth. Farmers are still making use of manual methods of grading fruit maturity and identifying disease ridden crops, which can result in poor grading due to weariness and misjudgment. The article proposes the use of region-based convolutional neural network, a deep learning algorithm for detection and classification of common dragon fruit diseases and detection of dragon fruit maturity. There are three key steps to this process of detection and classification. The first step includes identification of several regions of interest, which are the bounding box candidates using selective search. A broad convolutional neural network extracts features from each region in the second step. The last step performs grouping of support vector machine (SVM) that tries to classify the regions for the classification of different dragon fruit diseases and detecting its maturity. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

Detection and Classification of Diseases and Maturity of Dragon Fruits

The quick deployment of cloud with computing platforms has driven novel tendencies which shifted operations of networks. However, the cloud is facing several security issues and is susceptible because of suspicious tasks and attacks. This paper devises a new method to detect malicious activities in cloud. Here, first step is the simulation of cloud patterns, wherein the data outsourced by the users are utilized for detecting malicious behaviors. The data pre-processing is done to eradicate unnecessary data and noise contained in the data and is performed using a min–max normalization process. The selection of imperative features is done using distance measure, namely Hellinger distance for mining the essential features. The augmentation of data is performed to make the data appropriate for improved processing. The malicious behavior detection is performed by exploiting the Deep Q network wherein training is performed with Autoregressive chimp optimization algorithm (AChOA), which is developed by integrating chimp optimization algorithm (ChOA) and Conditional Autoregressive Value at risk (CAViaR). The proposed AChOA-based Deep Q network outperformed with the highest testing accuracy of 94%, sensitivity of 94.1%, and specificity of 91.4%. © 2022, The Author(s), under exclusive licence to Springer Nature Singapore Pte Ltd.

International Journal of Intelligent Robotics and Applications

An effective optimization enabled deep learning based Malicious behaviour detection in cloud computing

The cloud market is growing every day. So are cloud crimes. To investigate crimes that happen in a cloud environment, an investigation is carried out adhering to the court of law. Forensics investigations require evidence from the cloud. Evidence acquisition in the cloud requires formidable efforts because of physical inaccessibility and the lack of cloud forensics tools. Time is very crucial in any forensic investigation. If the evidence is preserved before the cloud forensic investigation, it can give the investigators a head start. To identify and preserve such potential evidence in the cloud, we propose a system with an artificial intelligence (AI)-based agent, equipped for binary classification that monitors and profiles the virtual machine (VM) from hypervisor level activities. The proposed system classifies and preserves evidence data generated in the cloud. The evidence repository module of the system uses a novel blockchain model approach to maintain the data provenance. The proposed system works at the hypervisor level, which makes it robust for anti-forensics techniques in the cloud. The proposed system identifies potential evidence reducing the effective storage space requirement of the evidence repository. Data provenance incorporated in the proposed system reduces trust dependencies on the cloud service provider (CSP). © 2022 Prasad Purnaye and Vrushali Kulkarni, published by De Gruyter.