Browse Items (11855 total)
Sort by:
-
Classification Algorithms Used in the Study of EEG-Based Epileptic Seizure Detection
Epilepsy is a neurological illness that has become more frequent around the world. Nearly 80% of epileptic seizure sufferers live in low- and middle-income nations. In persons with encephalopathy, the risk of dying prematurely is three times higher than in the general population. Three-quarters of people with brain illnesses in low-income countries do not receive the treatment they require. Recurrent seizures are a symptom of epilepsy, characterized by strange bursts of excess energy in mind. Experts agree that most people diagnosed with epilepsy may be managed successfully, provided the episodes are discovered early on. As a result, machine learning plays an essential role in seizure detection and diagnosis. Support Vector Machine(SVM), Extreme Gradient Boosting(Xgboost), Decision Tree Classifier, Linear Discriminant Analysis(LDA), Perceptron, Naive Bayes Classifier, k-Nearest Neighbor(k-NN), and Logistic Regression are eight of the most widely used machine learning classification algorithms used to classify EEG based mostly Epileptic Seizures. Almost all classifiers, according to the study, give an efficient process. Despite this, the results show that SVM is the most effective method for detecting epileptic seizures, with a 96.84% accuracy rate. For diagnosing Epileptic Seizures using EEG signals, the perceptron model has a lower accuracy of 76.21% percent. 2021 IEEE. -
An Intelligent System to Forecast COVID-19 Pandemic using Hybrid Neural Network
A current outbreak known as COVID-19 has been discovered from the coronavirus was informed by WHO. COVID-19 is a universal pandemic that has brought out the best and the worst of humanity. Due to an increase in the cases daily, COVID-19 is creating a menace to public health and establishes a disruption of the social and economic development of the countries. The problem is the hospitals are not able to provide proper facilities and treatments on time due to the lack of facilities in India. The purpose of this project to build an efficient hybrid deep learning model for forecasting the COVID-19 pandemic with multiple features that are responsible for the spread of COVID-19 in the top five states in India. In particular, a hybrid model that incorporates Auto-Regressive Integrated Moving Average and Long-term Short Memory is been used to forecast confirmed cases. The linear and non-linear dependencies in the dataset is been dealt with by an ARIMA-LSTM hybrid model. As a result, when compared to the outcomes of ARIMA, LSTM models independently, the hybrid model was giving better results and was performing well in forecasting COVID-19 cases. Through this, the policymakers will get prior information on COVID-19 cases in states which will help the government and healthcare departments to take prominent measures to prevent it. 2021 IEEE. -
Zero Trust-Based Adaptive Authentication using Composite Attribute Set
Rapid evolution of internet-oriented applications has increased the threats to confidential data. Single-factor authentication approaches are no longer sufficient to ensure user credibility. Multi-factor authentication schemes are also not tamper-proof. A Zero Trust, adaptive authentication-based approach that uses the user's past behavior can offer protection in this scenario. This paper proposes a system that collects a composite attribute set that includes the user behavior, attributes of the application through which the user is requesting access, and the device used. The enhanced collection allows the creation of detailed context that allows granular variance calculation and risk score. 2021 IEEE. All Rights Reserved. -
HydroIoT: An IoT and Edge Computing based Multi-Level Hydroponics System
The depleting area of cultivable lands is increasing demands for implementing improved techniques that could use less space and produce more than traditional farming. This situation is common in all the developing and under developed countries. With a motivation to contribute towards providing solution to this growing problem of food scarcity, a Multi-Level Hydroponics System is proposed. The proposed system combines best of all trending technologies like IoT, Edge Computing and Computer Vision and applies it to Hydroponics. A cultivation estimation system based on image processing is implemented and accuracy of the same is tested with actual produce. The crop used for the proposed system is corn as it serves as best fodder for cattle. It was observed that with proposed system up to 95% accuracy in estimating fodder produce was achieved. 2021 IEEE. -
A Relative Analysis on the Spotting of Cardiovascular Disease Employing Machine Learning Techniques
Heart is one of the significant segments in the human body since it powers blood to the all the pieces of the body. Blood courses through the vein. Cardiovascular sickness is corresponded with the blockage of vein. The sign of heart sickness depends whereupon condition is impacting an individual. The term coronary illness is ordinarily utilized instead of cardiovascular infection. Dilated cardiomyopathy, Heart failure, Arrhythmia, Pulmonary stenosis, Mitral regurgitation, Coronary artery disease, Myocardial infraction, Mitral valve prolapse, Hypertrophic cardiomyopathy are the sorts of coronary illness. The several machine learning techniques are analyzed to spot heart disease. This paper gives relative investigation of coronary illness expectation utilizing machine learning. 2021 IEEE. -
Application of AI in video games to improve game building
Video Games Industry has been welcoming AI like any other industry for various tasks, AI in gaming helps to convey a much more realistic gaming experience, amplify player interaction and satisfaction over extensive periods. Additionally, the gaming industry is utilizing Artificial Intelligence to liberate its staff by making game development automated, quicker, and less expensive. In this work an experiment is described using Deep Neural Network and Statistical techniques for forecasting the location of an object in future frames of a video, it focuses on the engineering phase of the game, the proposed model combines future prediction of object location which helps to build the infinite universe in the videogame without any additional videos frames of the input video or hard coding any scenes to build the scenes further. 2021 IEEE. -
Green and Sustainable Software Model for IT Enterprises
The present study is based on developing a Green and Sustainable software because in the present-day computing devices are used for all kinds of purposes and they consume a lot of energy to perform these services. The ICT sector itself consumes a lot of energy so there is a need to think of alternatives that can reduce the level of energy consumption, thus, green ICT practice can be a good option. There is, however, a scarcity of researches that explains how the maintenance of green knowledge in ICT software development may be implemented. Since we recognize that software development process (SDL) plays an essential role in enabling the ICT community, uncontrolled green knowledge in developing software that would lead to the dilemma of failing to satisfy both the community's business and environmental requirements. Therefore, this research will concentrate on presenting a methodology applying an innovative model for managing the green software development and implementation. Keeping this concern in mind the present paper is going to provide a Green and Sustainable software model which can be used in green ICT practices and will be helpful in reducing the energy consumption used by computers. 2021 IEEE. -
Distributed DoS Detection in IoT Networks Using Intelligent Machine Learning Algorithms
The threat of a Distributed Denial of Service (DDoS) attack on web-based services and applications is grave. It only takes a few minutes for one of these attacks to cripple these services, making them unavailable to anyone. The problem has further persisted with the widespread adoption of insecure Internet of Things (IoT) devices across the Internet. In addition, many currently used rule-based detection systems are weak points for attackers. We conducted a comparative analysis of ML algorithms to detect and classify DDoS attacks in this paper. These classifiers compare Nave Bayes with J48 and Random Forest with ZeroR ML as well as other machine learning algorithms. It was found that using the PCA method, the optimal number of features could be found. ML has been implemented with the help of the WEKA tool. 2021 IEEE. -
The effect of cutting fluid in improving the machinability of Inconel 718 using ceramic AS20 tool
Industries demand a vast usage of superalloys in heat resistant and high temperature applications. These include nozzle of rocket fuel engines, throttle valve of turbojet engines, turbine blade discs of aerospace industries, rivets and fittings of chemical and production industries, biomedical applications in super strength resistive steels. These superalloys such as Inconel 718 finds its vast applications in all such industries. To machine such materials a lot of wear and tear occurs at the cutting tool. Hence, cutting fluid helps in reduction of tool wear and improving surface roughness. In this paper, two cutting fluids Koolkut 40 and Hicut 590 have been used in emulsified form during the machining of Inconel 718 with Ceramic cutting tool. Hicut 590 has been seen a better heat resistive cutting fluid in reducing the tool wear and thus improving the life of the tool. 2021 Elsevier Ltd. All rights reserved. Selection and Peer-review under responsibility of the scientific committee of the Global Conference on Recent Advances in Sustainable Materials 2021. -
Alzheimer's Disease Detection using Machine Learning: A Review
Alzheimer's is a progressive brain disorder which is an untreatable, and inoperable and mostly affect the elderly people. There is a new case of Alzheimer's disease being discovered globally in every four seconds. The outcome is fatal, as it results in death. Timely identification of Alzheimer's disease can be beneficial for us to get necessary care and possibly even avert brain tissue damage by the time. Effective automated techniques are required for detecting Alzheimer's disease at very early stage. Researchers use a variety of novel approaches to classify Alzheimer's disease. machine learning, an AI branch use probabilistic technique that allow system to acquire knowledge from huge amount of data. In this paper we represent a analysis report of the work which is done by researcher in this field. Research has achieved quite promising prediction accuracies however they were evaluated the the non-existent datasets from various imaging modalities which makes it difficult to make the fair comparison with the other methods comparison among them. In this paper, we conducted a study on the effectiveness of using human brain MRI scans to detect Alzheimer's disease and ended with a future discussion of Alzheimer's research trends. 2021 IEEE. -
Application of Machine Learning in Customer Churn Prediction
Retaining customers is the central component of a company's growth strategy. It is evident that several industries are experiencing a surge in customer churn due to the global pandemic. As a result, customer retention that lies at the core of customer relationship management, has become the foundation for every industry to plan for future growth. By reducing customer churn, a company can maximize its profit. Studies suggest that significant advancements are made in the field of customer churn prediction in domains like telecom, banking, e-commerce and energy sector. The focus of the paper is to present a detailed review of the various machine learning techniques applied to address churn. Fifty-five papers related to churn classification published between 2004 and 2020 are collected and analyzed. The reviewed papers are categorized into five main themes. These themes are feature selection techniques, methods to handle class imbalance, experimentation with machine learning algorithms, hybrid models and ensemble models respectively. Finally, few suggestions are presented as direction for future research. 2021 IEEE. -
Co-MoS2 nanoflower coated carbon fabric as a flexible electrode for supercapacitor
Cobalt doped MoS2 (Co-MoS2) nanoflowers have been successfully synthesized via a simple one-step hydrothermal method for supercapacitor applications. To identify the crystalline nature and morphology, the as-prepared material is characterized by XRD, SEM, and TEM measurements. The material exhibits a specific capacitance value of 86 F g-1 at a current density of 1 Ag-1 in symmetric two-electrode configuration with excellent cyclic stability of 98.5% even after 10,000 chargedischarge cycles. The results suggest the suitability of Co-MoS2 as an efficient electrode material for supercapacitors. 2021 Elsevier Ltd. All rights reserved. -
A real time fog computing applications their privacy issues and solutions
Edge Computing (EC) has brought cloud technology to the channel's edge. It inherits some qualities from cloud services, but it also has some distinctive features such as geo-distribution, network connectivity, and reduced power. Along with the genetic inheritance, it also acquires the issues and concerns cloud computing services, such as renewable energy and resource allocation. This work provides a critical analysis of the fog architectural design in terms of security. Since 2018, the state of the artwork has been critically analyzed in terms of security mechanisms and security threats. The existing security methods are classified based on the security objectives they achieved. It would provide a complete and coherent difference between both the security areas investigated and those that have not. 2021 IEEE. -
Power Line Communication Parameters in Smart Grid for Different Power Transmission Lines
In an electrical power system smart grid is a network that renewable energy sources along with smart devices. Communication capabilities of the conventional grid can be improved by the inclusion of superior sensing and computing abilities. Device control, remote management, information collection, intelligent power management is achievable by using communication networks. Wired communication technology is used because of its advantages like reliable connection, free from interference, and faster speed. In this paper, the data communication parameters have been analyzed using Power Line Communication (PLC) with various lengths of transmission lines. An orthogonal Frequency Modulation scheme is used to obtain the minimum BER.MATLAB Programming has been carried out and the results have been compared with the standards and found to be satisfactory. 2021 IEEE. -
Rating of Online Courses: A Machine Learning Based Prediction Model
Online courses market has provided an economical and easy access to knowledge. When it comes to make a decision related to purchase of online course, little is known about what attributes can be depended upon to guess the quality of an online course. Ratings for online courses act as a reliable signal for assessing the quality of a course. The study discusses the prediction of ratings for online courses using Artificial Neural Network based on Particle Swarm Optimization (ANN-PSO). The experimental results suggests that ANN-PSO model has the capacity to predict the ratings for online courses on the basis of its attributes with accuracy. 2021 IEEE. -
A Comparative Analysis of Biodiesel Properties Derived from Meat Stall Wastes through Optimized Parameters
Biodiesel is considered as alternative green fuels that can be used in Internal Combustion engines as a replacement fuel for conventional diesel. Biodiesel is extracted from vegetable and animal sources which are rich in triglycerides. In this work, an attempt has been made to obtain and characterize the biodiesel from animal wastes such as chicken skin and pig tallow which are available in abundance and at an economical cost within the authors' geographical location. Initially, the feedstock is decontaminated and subjected to conventional heating to convert it into fatty oil. Heating is carried out at different temperatures and for varying time to find out the optimal combination of time and temperature, which would result in maximum fat yield. The fatty oil is then subjected to the trans-esterification process with methyl alcohol in the presence of a catalyst to extract crude biodiesel. A de-canter funnel is used to separate the glycerine and biodiesel from the crude extract. The extracted biodiesel is mixed in different volume percentages with conventional diesel, and various thermochemical properties were evaluated as per ASTM standards. The test result indicated that the properties of the biodiesel blends were well within the limits as prescribed by ASTM standards. Published under licence by IOP Publishing Ltd. -
Analysis of benchmark image pre-processing techniques for coronary angiogram images
Coronary Artery supplies oxygenated blood and nutrients to the heart muscles. It can be narrow by the plaque deposited on the artery wall. Cardiologists and radiologists diagnose the disease through visual inspection based on x-ray images. It is a challenging part for them to identify the plaque in the artery in the given imagery. By using image processing and pattern recognition techniques, a narrowed artery can be identified. In this paper, pre-processing methods of image processing are discussed with respect to coronary angiogram image(s). In general the angiogram images are affected by device generated noise / artifacts; pre-processing techniques help to reduce the noise in the image and to enhance the quality of the image so that the region of interest is sensed. The main objective of the medical image analysis is to localize the region of interest by removing the noise. It is essential to find the structure of the artery in the angiogram image, for that preprocessing is useful. 2021 IEEE. -
Envisioning the potential of Natural Language Processing (NLP) in Health Care Management
Natural Language Processing (NLP) continues to play a strategic role in disease detection, intensive care, drug discovery and control of mushrooming infections during the current pandemic. It energizes chat programs to reduce outbreaks during the initial stages of coronavirus infection. NLP technologies have reached new heights in terms of utility, and are at the heart of the success of a multilingual conversation system, and Deep learning language models. It supports more languages around the world. NLP powered AI such as Health map and Copweb platforms track patient requests and perform incident detections. This study looks at the role of NLP and its technologies, challenges, and future possibilities using AI and machine learning for crisis mitigation and easier electronic health records (EHRs) maintenance in the health care industry. This research work explores the strategic approach and potential of NLP which maximizes the value of the EHR and healthcare data, making data a critical and trusted component in improving health outcomes 2021 IEEE. -
ML based sign language recognition system
This paper reviews different steps in an automated sign language recognition (SLR) system. Developing a system that can read and interpret a sign must be trained using a large dataset and the best algorithm. As a basic SLR system, an isolated recognition model is developed. The model is based on vision-based isolated hand gesture detection and recognition. Assessment of ML-based SLR model was conducted with the help of 4 candidates under a controlled environment. The model made use of a convex hull for feature extraction and KNN for classification. The model yielded 65% accuracy. 2021 IEEE. -
Experimental Investigation of Air Circulation Using Duct System in a Non-AC Bus Coach
Public transport is the life line in many of the developing and under developed countries for the safe conveyance, i.e. also consider as economical. The major limitation in public transport (non-AC busses) Air Condition, is the lack of proper air circulation leading to suffocation and vomiting. The present research work emphasis on design and analysis of air flow duct system (non AC Busses) to increase the level of comfortance of the passengers, tools like solidworks software 2016 is used for 3D drawing, Hypermesh software 13.0 is for the discretization and ANSYS Fluent software 16.0 for the Computational Fluid Dynamic (CFD) analysis, from the experimental the airflow is found to be 10 m/s, and from the numerical analysis the airflow is found to be 9.8 m/s, by comparing the experimental and numerical results a negligible deviation of 2% is observed and it is within the limit. Published under licence by IOP Publishing Ltd.