Volume 9 | Issue 2

 Abstract

Evolution of Credit Scoring Models in Indian Banking - An Empirical View

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Impact of MGNREGA on Labour Relations and Budget Constraints - An Analysis

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

In today�s world information storing and retrieval plays an important role. Database systems play a key role in the new commercial system for information storage. For accessing the data from database, a person who is not having the knowledge of SQL may find themselves handicapped while dealing with the database. This presents a significant limitation in a developing country such as India, because even today, a very large majority of the population does not have the technical knowledge of how to deal with database systems. In such a case, Natural language processing (NLP) assumes a significant job. NLP is getting perhaps the most dynamic strategies utilized in Human-PC Interaction. It is a part of AI which is utilized for Information Retrieval, Machine interpretation, and Language Analysis. This paper, proposes a system which allows the user to query the database in a compatible mode language, through a convenient Graphical User Interface which results in the data required by the user.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Natural language processing, Database, Artificial intelligence, Information retrieval.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

It is anticipated that 40% of all of the meals manufacturing made in India are wasted as meals leftover. Out of the whole meals, leftover wasted yearly, 23% comes with inside the shape of eateries leftovers, and 7% comes in the form of stores leftovers [1]. The quantity of meals wasted from eating places and meals stores is enormous, and if, somehow, the meals this is being wasted in those eating places and meals chains are furnished to the needy, then it could be a supply of meals for lakhs of human beings daily. The key concept in the back of the idea is to digitalize the method of leftover meals management. Currently, the leftover meals from meals stores more often than not finally end up in landfills, which becomes a supply of world warming gases emission.[2] To upload to the problem, if the eating place attempts to donate the leftover meals to the needy, it has to take in the duty of transportation and dispensing. On the opposite hand, NGOs which can be operating difficult to offer meals to the terrible and the needy, locate it difficult to return back with the aid of using resources. We recommend creating software so as to act as a bridge between the NGOs and the eating places. We motive to contain the not unusual place human beings withinside the method of imparting meals to the needy with the aid of using imparting a software to the customers as well, in which they are able to donate any quantity to the particular eating place, and the eating place in going back will offer the meals same to the quantity donated.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Food Management, AWS Cognito, AWS SQS, Android Application, Leftover Food

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Cloud computing offers a flexible and convenient means of exchanging data, with several advantages for both society and individuals. With the widespread usage of cloud emails and frequent reports of large-scale email leakage occurrences, the security property known as forward secrecy has become desirable and necessary for both users and cloud email service providers to increase the security of cloud email systems. Typical techniques of attaining forward secrecy, such as Diffie-Hellman key exchange and forward-secure public-key encryption, have not been extensively authorised and used because they fail to meet the security and practicality requirements of email systems at the same time. We introduce a new cryptographic primitive called forward-secure puncturable identity-based encryption (fs-PIBE) in this paper to capture forward secrecy of encrypted cloud email systems without sacrificing practicability. It allows an email user to perform fine-grained decryption capacity revocation. In the standard model, we design a framework for encrypted cloud email systems and instantiate it with a concrete fs-PIBE structure that has constant ciphertext size and proved security. We enhance the proposed fs-PIBE scheme to provide end-to-end encryption and outsourced decryption, respectively, to improve the security and efficiency of the presented framework.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Secure Cloud Email, Identity-based Encryption, Broadcast Encryption, Encrypted cloud emails, forward secrecy, puncturable encryption, identity-based encryption, end-to-end security.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

In the fields of healthcare and bioinformatics, the definition of breast cancer has been the topic of concern, since this is the second principal explanation for death from cancer in women. This form of study will only be achieved if a sample of breast tissue is extracted from the breast, evaluated, and analyzed under a microscope. In the histopathology lab, problems are detected by examining the specimens using trained and qualified pathologists, and identified with further investigation using special techniques. However, the ultrasound can incorrectly identify pathological changes or conditions because they have practiced in this specialty before. Several pattern recognition studies have recently identified a lot of areas for potential for improvement, and, hence there is now an increased focus on constructing powerful image processing experiments to create a highly-and improving existing diagnoses. Let�s use the histology and image recognition techniques to identify the disease types of breast through the use of the image feature extraction technique and deep learning approaches for the image feature. This picture can be expanded using ultrasound processing and CNN technique is available before ultrasoundfeature extraction is applied and the final classification in ultrasound feature extraction.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

image classification, breast cancer diagnose, feature extraction, classification,deep Learning.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Confidentiality, integrity and auditability have been studied for long in cryptography, and to provide security is the foundation for all modern cryptographic research. In next generation of cloud computing client store data on centralized cloud server may some challenges. This work is focusing on honesty of data stored in cloud data servers. Cloud storage is one of the service provided by Cloud computing in which data is maintained, managed, backed up remotely andmade available to users over a network (typically the Internet). The user is concerned about the integrity of data stored in the cloud as the user�s data can be attacked or modified by outside attacker. Therefore, a new concept called data auditing is introduced which check the integrity of data with the help of an entity called Third Party Auditor (TPA). The purpose of this work is to develop an auditing scheme which is secure, efficient to use and possess the capabilities such as privacy preserving, public auditing, maintaining the data integrity along with confidentiality. Thus the new auditing scheme has been developed by considering all these requirements. It consist of three entities: data owner, TPA and cloud server. The data owner performs various operations such as splitting the file to blocks, encrypting them, generating a hash value for each, concatenating it and generating a signature on it. The TPA performs the main role of data integrity check. It performs activities like generating hash value for encrypted blocks received from cloud server, concatenating them and generates signature on it. It later compares both the signatures to verify whether the data stored on cloud is tampered or not. It verifies the integrity of data on demand of the users. The cloud server is used only to save the encrypted blocks of data.The data integrity verification is done by using a TPA which is allow to check integrity of data periodically on rather than client. When data integrity is lost then client gets information from TPA. Not only checking of data purity also supports data dynamics. In this auditing task observes data insertions, deletions and modifications. This system is skillful of supporting both public auditability and data dynamics. Markel Hash Tree is used to improve block level authentication. In order to handle auditing tasks simultaneously, bilinear aggregate signature is used. This permits TPA to execute auditing simultaneously for many clients. In this dissertation the evaluation of multi user based TPA system is presented.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Cloud Storage; TPA; Privacy Preserving ; Public Auditing; Integrity, Third party Auditor (TPA), Cloud storage Server (CSS), Cloud service provider (CSP), Proof of Retrievability (PoR), Provable data Possession (PDP), Markel Hash Tree (MHT).

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Research in forestry is one of the areas where drones have recently found application, thanks to the high-resolution data that can be obtained in a short period of time and at a low cost. While manned air-craft and satellite remote sensing have their place, drones can fill in many of the gaps left by these methods, particularly when it comes to forest management and land use. The goal of this research is to provide a brief overview of the many UAV (Unmanned Aerial Vehicle) applications in forestry, such as forest mapping, forest management planning, canopy height model generation, and mapping forest gaps. This study. They have a wide range of potential applications in the near future and should be implemented quickly in a variety of settings for forest management sustainability.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Drones, UAV, remote sensing, forest management.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

70% of the Earth�s surface is covered by water, but only 0.3% of total water is potable. Rapidly growing population and industrialization have raised serious problems for the world. It is difficult to clean polluted waters, because of the many problems faced by applications of time-consuming traditional treatment methods. The electrochemical coagulation method comes to rescue with its exceptional efficiency. Electrochemical coagulation technology is a method of treating wastewater using electricity. Which makes it more effective and economical than traditional methods. This advanced method has the advantage of removing the smallest colloidal particles compared with conventional activated sludge process and other technologies. Along with this, the electrocoagulation process is flexible according to the intensity of impurities present in wastewater by changing the parameters affecting it. This review paper reviews and discusses the mechanism of electrocoagulation process, and discusses the parameters affecting the rate of electrocoagulation efficiency along with the rate of impurity removal by electrodes in different conditions.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Wastewater, Electrocoagulation, Electrodes, Adsorbent, Coagulant, Treatment, Impurities, BOD, COD.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

GLAUCOMA is an irreversible and incurable eye infection in which the optic nerve is consistently hurt. It leads to slow progressive degeneration of the retina. Though it can not be fully cured, the progression can be controlled by detection in the early stage. Glaucoma affects the Optic Head Nerve and its density is dependent on measurement of Optic Cup and Optic Disc. We can develop a model in which glaucoma can be recognized/detected in the starting period only so that further loss of vision can be prevented. We can use different algorithms like CNN, logistic regression, K- nearest neighbors classifier, Support Vector machines. All these techniques can be used to measure OC and OD dimension. We are planning to use different datasets like ARIA, RIGA and will be using FUNDUS images which later will be converted into segmented images. Our aim will be to provide a comprehensive overview of various machine learning algorithms which are used to detect GLAUCOMA based on fundus images and finding out the best algorithm which gives the best accuracy

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Glaucoma, Intraocular Pressure(IoP), ResNet, LAG, ARIA, VGG.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Breast cancer is the most common type of cancer among Indian women. One in every two Indian women diagnosed with breast cancer dies, resulting in a fifty percent likelihood of death. The goal of this work is to compare three widely used machine learning algorithms and methodologies for breast cancer prediction: Random Forest, kNN (k-Nearest-Neighbor), and Nave Bayes. The Wisconsin Diagnosis Breast Cancer data set was used as a training set to examine the performance of several machine learning approaches in terms of accuracy and precision. Iris melanocytic tumours, often known as eye tumours, are the most dangerous tumours in the eye. Freckle, nevus, melanocytoma, Lisch nodule, and melanoma are all examples of this. Early detection of an eye tumour is quite challenging. Many studies are being conducted to detect eye problems. However, just a few studies on ocular tumours have been published. To capture the region, the majority of the system requires specific data gathering equipment. This is quite costly. Doctors recommend PET - CT, eye ultrasound, angiography, optical coherence tomography, and other tests to diagnose eye melanoma. Using a deep learning technique, a new method for detecting eye tumours from eye pictures is given. Though current medical imaging science is advancing at a breakneck pace, pinpointing a brain tumour remains a difficult endeavour. Unlike other imaging systems, medical imaging has the largest penalty for a minor inaccuracy. To reduce the risk of inaccuracy, tumour detection should be precise. Previous studies relied on biopsy to distinguish tumour tissue from other soft tissues in the brain, which is time-consuming and prone to error. Melanoma is the most lethal type of skin cancer in the globe. Many efforts have been made to use deep learning to detect melanoma early using dermoscopic pictures. For a correct diagnosis of melanoma, it is critical to identify certain lesion patterns. The prevalent lesion patterns, on the other hand, are not always present, resulting in sparse labelling issues in the data. We present a multi-task U-Net model to detect melanoma lesion properties automatically in this work.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

multi disease detection, pre-processing, classifier algorithm, feature extraction, Convolutional neural network (CNN) etc.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

For recognizing the face of any person, the most prime feature is the face, face recognition method is the best method for recognizing an individual than any other biometric methods. In face recognition, system can identify a person from video or image but in this proposed system, the face is identified only from real-time images. The process for authenticating face is partitioned into two steps, first step is to detect face using Haar Cascade algorithm which is a machine learning object detection algorithm, it does object detection using haar feature based cascade classifiers which is an effective object detection method purposed by Paul Viola & Michael Jones. The second step is to recognize an individual�s face using LBPH (Local Binary Patterns Histogram) algorithm. The main purpose of this research is to recognize and detect an individual�s face in real-time camera-based image capturing.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

LBPH(Local Binary Pattern Histogram) Algorithm, Haar Cascade Algorithm, OTP.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Speech Emotion Recognition (SER) is a term used to describe a system that identifies the human emotion and related feeling states from the speech. This is aided by the evidence that the speech frequently expresses basic emotion through pitch and tone. In recent years, emotion identification has become a rapidly growing research subject. Machines, in contrast to individuals, lack the ability to see and express emotions. However, by using automated emotion recognition, human-computer interaction can be enhanced, eliminating the need for human assistance. Emotional speech signals are used to examine basic emotions such as disgust, calm, fear, happy, and so on. We focused on studying Human Speech Emotion using a feature extraction technique called Mel Frequency Cepstrum Coefficients (MFCC), as well as anticipating emotions using three different classifiers: Convolutional Neural Network (CNN), Support Vector Machine (SVM), and Multilayer Perceptron (MLP). It has been discovered that every one of these strategies gives the best accuracy with a specific arrangement of features separated from the input voice. This paper analyzes the accuracy of SVM, CNN, and MLP with their own set of features.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Speech emotion recognition, mel cepstral coefficient, convolutional neural network, multi-layer perceptron, support vector machine.

  License

Creative Commons Attribution 4.0 and The Open Definition