Browse Items (2150 total)
Sort by:
-
Classification of Disaster Tweets using Machine Learning and Deep Learning Techniques
Social networks provide a plethora of information for gathering extra data on people's behavior, trends, opinions, and feelings during human-affecting occurrences, such as natural catastrophes. Twitter is an inevitable communication medium during calamities. People mainly depend on Twitter to announce real-time emergencies. However, it is rarely straightforward if someone is declaring a tragedy. Sentiment analysis of disaster tweets aid in situational awareness and realizing the disaster dynamics. In our paper, we perform a sentimental analysis of disaster tweets using techniques based on machine learning and deep learning. The tweets are pre-processed before being converted into a structured form using Natural Language Processing (NLP) methods. Supervised learning techniques such as the Support Vector Machine and the Naive Bayes Classifier algorithm are used to develop the Classifier, which categorizes tweets into distinct catastrophes and selects the most appropriate algorithm. The chosen algorithm is further enriched with an emoticon detection algorithm for explicit elucidation. Our research would help disaster relief organizations and news agencies to conclude about the state of affairs and do the needful. 2022 IEEE. -
Machine Learningcloud-Based Approach to Identify and Classify Disease
The term "Internet of Things"(IoT) describes the process of creating and modeling web-related physical objects across computing systems. IoT-based healthcare applications have offered multiple real-time products and benefits in recent years. For millions of people, these programmers provide hospitalization can get regular medical records and healthy lives. The introduction of IoT devices in the health sector has several technological developments. This study uses the IoT to construct a disease diagnostic system. Wearable sensors in this system initially monitor the patient's sympathy impulses. The impulses are then sent by a network environment to a server. In addition, a new hybrid approach to evaluation decision-making was presented as part of this research. This technique starts with the development of a set of features of the patient's pulses. Based on a learning approach qualifications are neglected. A fuzzy neural model was used as a diagnostic tool. A specific diagnosis of a particular ailment, such as the diagnosis of a patient's normal and abnormal pulse or the assessment of insulin issues, would be modeled to assess this technology. 2022 IEEE. -
Patient Monitoring System for Elderly Care using AI Robot
The use of robots in numerous industries has expanded in recent decades. Self-guiding robots have started to arise in human life, particularly in sectors pertaining to the lives of old people. Age-related population growth is accelerating globally. As a result, there is a rising need for personal care robots. The purpose of this requirement is to increase opportunities for mobility and support independence. To meet this demand, a robot with specific functionalities to help older people has been designed. The standard values of healthcare parameters are stored in the database by recording and comparing the current values the system will give an alarm and also sends a message to the doctor or caretaker so that a proper care would be given to the patients. We are including a preset distance value to monitor the elder people. Here we are using some sensors to detect the health parameters from the person. Robot have designed to intimate the family members if any changes occur in the health parameters. It helps the people to stay alone in home with safe manner. 2022 IEEE. -
Comparison of Augmentation and Preprocessing Techniques for Improved Generalization Performance in Deep Learning based Chest X-Ray Classification
Convolutional Neural Network (CNN) models are well known for image classification; however, the downside of CNN is the ineffectiveness to generalize and inclination towards over-fitting in case of a small train dataset. A balanced and sufficient data is thus essential to effectively train a CNN model, but this is not always possible, especially in the case of medical imaging data, as often patients with the same disease are not always available. Image augmentation addresses the given issue by creating new data points artificially with slight modifications. This study, investigates ten different methods with various parameters and probability and their combined effect on the test dataset's generalization performance and F1 Score. For the study, three pre-Trained CNN models, namely ResNetl8, ResNet34, and ResNet50, are fine tuned on a small training dataset of 500 Pneumonia and 160 Non-Pneumonia(Normal) Images for each augmentation setting. The test accuracy, F1 Score, and generalization performance were calculated for a test dataset consisting of 50 Pneumonia and 16 Non-Pneumonia(Normal) Images. 2022 IEEE. -
Academic Certificate Validation Using Blockchain Technology
Academic certificates are essential for an individual's career and hence they are more prone to being tampered. This paper proposes an idea of sharing certificates and verifying their authenticity using blockchain technology. Blockchain paves the way for secure storage and sharing of information. Its main focus is to maintain trust among users. This proposal focuses on designing and implementing a system that will prove to be a solution for addressing the issue of fake certificates using Hyperledger Fabric. The technology here is tamper-proof and maintains transparency. This system will have a database of academic certificates awarded by the University, which is recorded as a transaction using the Hyperledger Fabric, which further can be referred by other organizations present in the network to verify the authenticity of the certificates using the information provided by the students to the database. This system provides end to end encryption. 2022 IEEE. -
Classification and analysis of Alzheimer's Disease using Deep Learning methods on MRI and PET
Alzheimer's disease (AD) falls in the category of neurodegenerative illness in which an individual loses his or her power to remember things and behaviors. It affects memory in younger patients and as it progresses causes diffuse cortical functions. However, a major issue with the diagnosis and treatment of AD symptoms is that it has complex pathogenesis because of which there is no clinical intervention for its treatment. There is no disease-modifying treatment to cure AD symptoms that increases co-morbidities among the patients. The present research identified this gap and focuses on using Deep Learning methods on MRI and PET data so that there is early diagnosis of AD by healthcare experts and they could propose a better treatment process for reducing AD symptoms. The present research identified that by using deep learning-based approaches particularly ResNet50 architecture, there is the execution of quantitative assessment of brain MRI and PET to acquire insights about the internal abnormalities through self-learning features. It will help in initiating proper treatment and avoiding damage to the brain further. 2022 IEEE. -
Automatic Resume Parsing using Greywolf Algorithm Integrated with Strategically Constructed Semantic Skill Ontologies
The quest for finding the right candidate for their post has made the recruiters employ several methods since the beginning of corporate job recruitment. Apart from the skills and the quality of interview, a thing that matters the most and forms the basis of selection is the candidate's resume. Recruiters and companies have a tough time dealing with the several thousands resumes of the candidates which apply, as manually scanning them and finding the right selection can be tough most of the time. In this paper, Natural Language Processing(NLP) methods have been integrated with ontologies to improve the pace and quality of the recruitment process by proposing an automatic resume parser model. The resume of a candidate, along with his LinkedIn and GitHub profiles are weighted and using the Greywolf algorithm, the global maxima of the most deserving and qualified candidate are found and are recommended with a high accuracy of 96.13%. 2020 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0) -
Kidney Abnormalities Detection and Classification Using CNN-based Feature Extraction
The presents of noises degrade the quality of ultrasound images and diminishes the disease diagnosis accuracy. Thus, an effective automatic stone and cyst detection system is beneficial to both the medical practitioners and patients. In this paper, an automatic detection and classification system for kidney stone and cyst image is proposed. The Gaussian filtering and Contrast Limited Adaptive Histogram Equalization (CLHE) techniques are applied to improve the quality of the images. In the next step, segmentation has been done based on the entropy of the image. The gamma correction technique has been applied to improve the overall brightness and an optimal global threshold value is selected to extract the region. The CNN model has attained much attention in medical image recognition and classification. In this paper, the pre-trained model ResNet-50 is utilized as a feature-extractor and Support Vector Machine as classifier to categorize the normal, cyst and stone images. The CNN model is analyzed with various other classification models such as k-nearest neighbor, decision tree and Nae Bayes. The results demonstrate that the ResNet-50 with supervised classification algorithm SVM is an optimal solution for analyzing kidney diseases. 2022 IEEE. -
Descriptive Answer Evaluation using NLP Processes Integrated with Strategically Constructed Semantic Skill Ontologies
The world is moving towards an online methodology of education. One of the key challenges is the assessment of questions which do not have a definite answer and have several correct answers. To solve this problem, and for quality evaluation of descriptive answers online, an automatic evaluation methodology is proposed in this work. A language model is modelled from the expected answer key, and entity graphs are generated from the ontology modelled using the input answer to be evaluated. Natural Language Processing (NLP) techniques like Stemming, Summarization, and Polarity Analysis are integrated in this work with Ontologies for the efficient evaluation of descriptive answers. Several challenges which come across evaluating descriptive answers are discussed in this chapter, and they have been solved in order to obtain a dynamic and robust evaluating system. Finally, the system is evaluated using a user-feedback methodology comprising a panel of 100 students and 100 professors. 2022 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0) -
Efficiency Enhancement using Least Significant Bits Method in Image Steganography
Over the years, there has been a tremendous growth in the field of steganography. Steganography is a technique of hidden message passing i.e. transferring a message which is not visible to human eyes, through some media such as an image, music, games etc. In this particular article we focus on Image steganography which has its own advantages and has undergone a lot of improvements in the past years. The most basic image steganography can be achieved by changing the LSBs (Least Significant Bits) of the image pixels. These bits can usually be called the redundant bits. However, changing a large numbers of LSBs of an image can distort the image to an extent where it would be easily noticeable that the image maybe carrying a hidden message rendering it useless. These LSBs are changed according to the message bits allowing the person to hide their data which can be decoded later by reading the LSBs of image pixels. This paper introduces and explains a method to improve the efficiency of LSB method. 2022 IEEE -
A Slotted Circular Patch Antenna with Defected Ground for Sub 6 GHz 5G Communications
In this paper, a slotted circular patch antenna with Defected Ground Structure (DGS) is presented. The slots created on radiating element and the defect introduced on the ground plane shifted the resonance frequency from 2.49 GHz to 1.17 GHz. This corresponds to 53% reduction in size at 1.17 GHz. The proposed antenna is designed on FR-4 substrate (r=4.4) with thickness of 1.6 mm. Simulations are carried out using HFSS Ver. 18.2. The simulated reflection coefficient of Circular Patch Antenna (CPA) at 2.49 GHz, Slotted Circular Patch antenna (SCPA) at 2.34 GHz and Slotted Circular Patch antenna with Defected Ground Structure (SCPA-DGS) at 1.17 GHz are - 28.7 dB, -31.33 dB and -11.03 dB respectively. For validating the simulated design, SCPA-DGS is fabricated and measured its reflection coefficient and VSWR using Vector Network Analyzer (Anritrsu S820E). The measured and simulated values are very well matched with each other. Therefore the proposed antennas may be used in sub 6 GHz 5G communication applications. 2022 IEEE. -
Quantum Information Processing for Legal Applications through Bloch Sphere of Law
The objective of the research work is to propose a quantum information processing model (QIP) for legal applications including litigation and investigation phases. The quantum information processing and quantum computing concepts can be visualized within a Bloch Sphere of Law (BSL) as legal Bloch vectors (LBV) as quantum computing entities. This quantum approach is needed since the complexity of legalities and the legal objects involved in the final judgement are to be reversible with a lot of uncertainties. The reasoning and prosecution through various trials and investigations are to be considered as mathematical matrix or unitary operations in this muti dimensional legal space. The mapping of legal information into technical and then vectorial representations are deployed through a glossary of legal terms in this quantum paradigm. As a forerunning study and application in the quantum paradigm, mathematical and computational models have been proposed in the work with a case study of a recent civil case. 2022 IEEE. -
Scalable, Cost Effective IoT Based Medical Oxygen Monitoring System for Resource Constrained Hospital Environment
Oxygen therapy is one of the critical treatments employed in epidemics, pandemics, and natural calamities. Recent covid pandemic worldwide witnessed many deaths due to improper management, delayed delivery, and wastage of medical oxygen. Therefore, efficient utilization of available oxygen is very important. To monitor and manage oxygen, several hospitals employ IoT-based systems. Scalability is an essential feature in such monitoring systems in order to cater to the needs of a sudden surge in the number of patients requiring oxygen. The most commonly employed technique to monitor and manage an oxygen cylinder uses a pressure sensor where scaling up is an issue. Therefore, in this paper, a scalable solution that efficiently measures and monitors the available oxygen in the cylinder is proposed. The approach measures oxygen level using a weight sensor module and raises alerts during critical conditions such as low oxygen level and blockage or leakage of oxygen. The proposed system is a cost-effective, plug-and-play system that aids rapid deployment thereby providing timely care to the patients. Also, it does not require any change in the existing infrastructure making it suitable for a resource-constrained environment. The proposed system supports a web-based dashboard and mobile app that can be remotely accessed. 2022 IEEE. -
IRIS Data Classification using Genetic Algorithm Tuned Random Forest Classification
Optimising hyper-parameters in Random Forest is a time-consuming undertaking for several academics as well as professionals. To acquire greater performance hyper-parameters, specialists should explicitly customize a series of hyper-parameter settings. The best outcomes from this manual setting are then modelled and implemented in a random forest algorithm. Several datasets, on the other side, need various prototypes or hyper-parameter combinations, which may be time-consuming. To solve this, we offered various machine learning models and classifiers for correctly optimising hyper-parameters. Both genetic algorithm-based random forest and randomised CV random forest were assessed on performance measures such as sensitivity, accuracy, specificity, and F1-score. Finally, when compared to randomised CV random forest, our suggested model genetic algorithm-based random forest delivers more incredible accuracy. 2022 IEEE. -
Trident Shaped Compact Planar Antenna for Microwave Applications
A compact planar antenna for X/Ku-band microwave communication is suggested in this paper. The presented geometry is capable of radiating the large frequency band from 6.8 to 20GHz, which covers the X-Band/Ku-Band Communication with high efficiency. The impedance bandwidth of the radiator is 98.5%, with an electrical size of. 34?x.34?x0.034A in lambda. The suggested design includes a modified patch in the trident shape fed by a microstrip line. Rectangular elements have been designed for better resonances at lower modes. The antenna is simulated with an FR4 substrate using CST Simulator. The exact dimensions of the antenna are 15x15x1.5 cubic millimeter. Five stages evolution process is also investigated by simulations, and corresponding S-parameter results are presented. The proposed structure also demonstrates stable radiation patterns across the operating bandwidth. The proposed radiator has a high gain of 3.1 dBi, and an efficiency of 87%. Therefore, it is useful for X-band, and Ku-band, including Radar, Space, Terrestrial, and Satellite microwave communication. 2022 IEEE. -
A JSON Web Signature Based Adaptive Authentication Modality for Healthcare Applications
In the era of fast internet-centric systems, the importance of security cannot be stressed more. However, stringent and multiple layers of security measures tend to be a hindrance to usability. This even prompts users to bypass multi-factor authentication schemes recommended by enterprises. The need to balance security and usability gave rise to Adaptive authentication. This system of utilizing the user's behavioral context and earlier access patterns is gaining popularity. Continuously analyzing the user's request patterns and attributes against an established contextual profile helps maintain security while challenging the user only when required. This paper proposes an Open standards based authentication modality that can seamlessly integrate with an Adaptive Authentication system. The proposed authentication modality uses JavaScript Object Notation(JSON), JSON Web Signature(JWS) and supports a means of verifying the authenticity of the requesting client. The proposed authentication modality has been formally verified using Scyther and all the claims have been validated. 2022 IEEE. -
Anticounterfeiting Method for Drugs Using Synthetic DNA Cryptography
Counterfeited products are a significant problem in both developed and developing countries and has become more critical as an aftermath of COVID-19, exclusively for drugs and medical equipment's. In this paper, an innovative approach is proposed to resist counterfeiting which is based on the principles of Synthetic DNA. The proposed encryption approach has employed the distinctive features of synthetic DNA in amalgamation with DNA encryption to provide information security and functions as an anticounterfeiting method that ensures usability. The scheme's security analysis and proof of concept are detailed. Scyther is used to carry out the formal analysis of the scheme, and all of the modeled assertions are verified without any attacks. 2022 IEEE. -
Comparative Study Analysis on News Articles Categorization using LSA and NMF Approaches
Due to exponentially growing news articles every day, most of their important data goes unnoticed. It is important to come up with the ability to automatically analyse these articles and segregate them based on the context and related to their particular domain. This paper applies topic modelling which is one of the most growing unsupervised machine learning fields on a million headlines articles in order to produce topics to describe the context of the news article. There are various generative models but we specifically focusing on the non-negative matrix factorization (NMF) and Latent Semantic Analysis (LSA) for implementing and evaluating news dataset. Furthermore, the findings reveal that both NMF and LSA are useful topic modelling tools and classification frameworks, but based on the experimental results the LSA model performed well to identify the hidden data with better mean coherence values and also consumes lesser execution time than NMF. 2022 IEEE. -
A Performance Investigations of Modular Multilevel Inverter with Reduced Switch Count
A multilevel inverter is a special variant of converter for dc-Ac conversion in medium and high voltage and power requirements. In this paper, a novel configuration with fewer switches needed has been developed for the staircase output voltage levels. Two direct current voltage sources and eight transistors are required to synthesize five levels across the load using the conventional topology. The modular topology has two dc voltage sources, and six switches with a five-level output. Using the optimum multi-carrier pulse width modulation approach, the voltage quality is enhanced and total harmonic distortion is reduced. Furthermore, the viability of the proposed topology in contrast to the conventional cascaded H-bridged multilevel inverter with five levels is established by presenting comparable results showing reduced power losses with varied modulation indexes and increased efficiency. The simulation analysis has been carried out using the MATLAB/SIMULINK tool. 2022 IEEE. -
Real-time Litter Recognition Using Improved YOLOv4 Tiny Algorithm
Littered roads have become a familiar sight in India. The main reason is the increasing population and inefficient waste disposal system. Since garbage collectors cannot pick litter in all the places, there is a need for an efficient way to detect it. Hence, a machine learning-based object detection model is used. In this, we have applied an improved YOLOv4-Tiny algorithm to detect the garbage, classify it and make the detection process easier on custom datasets. We have improved the algorithm in terms of the object prediction time, this is done by replacing a max pooling layer with one of two layers present in a fully connected layer. When an input is given, the algorithm detects the litter in the image with a bounding box around it along with the label and confidence score. The proposed model reduces the prediction time by 0.517 milliseconds less than the original algorithm employed which concludes that the object is predicted faster. 2022 IEEE.