Browse Items (2150 total)
Sort by:
-
Priority-driven Unbalanced Transportation Problem (PUTP) to obtain better Initial Feasible Solution
In this paper, we tackle the Priority-driven Unbalanced Transportation Problem (PUTP), a scenario where total demand exceeds total supply. An innovative algorithm, the Penalty-driven Priority-driven Unbalanced Transportation Problem (PPUTP) is introduced to solve this challenge. PPUTP allocates supplies to high-priority demands by computing penalties and sequentially addressing the most penalized demands, thereby ensuring priority demands are met efficiently. A comparative analysis with Vogel's Approximation Method (VAM) across various problem sets ranging from 5x5 to 50x50 dimensions demonstrates the efficiency of our algorithms. PPUTP consistently shows lower percentage increments from the optimal solution, indicating its robustness in providing near-optimal solutions. This study highlights the importance of algorithm selection based on problem set dimensions and complexity in Priority-driven Unbalanced Transportation Problem, with PPUTP emerging as a versatile and robust solution across various scenarios. 2024 IEEE. -
Rating-Based Cyberbullying Detection with Text, Emojis on Social Media
In the dynamic landscape of online interactions, cyberbullying has become pervasive, profoundly impacting user's digital well-being. Public figures, especially celebrities and influencers, face heightened vulnerability to online harassment, exacerbated by the post-pandemic surge in social media usage. To address this challenge, our research adopts a holistic approach to detect cyberbullying in text, considering both textual content and the nuanced expressions conveyed through emojis on social media platforms. We employed a diverse set of machine learning and deep learning models, including Support Vector Classifier, Logistic Regression, Random Forest, XGBoost, LSTM, Bi-LSTM, GRU, and Bi-GRU, to accurately classify non cyberbullying or cyberbullying text. Beyond classification, our study introduces an offensive rating system, assigning severity ratings on a 1-5 scale to identify cyberbullying instances. A critical aspect is the establishment of a threshold value which depends on user security and safety ethics of different social media platforms; texts surpassing this trigger an automatic recommendation to block the user, ensuring a proactive response to minimize harm. This recent contribution not only comprehensively addresses cyberbullying but also empowers society. 2024 IEEE. -
Optical Character Recognition system with Projection Profile based segmentation and Deep Learning Techniques
Optical character recognition is the solution to convert text from printed or scanned documents into editable data. This project is aimed at building a Optical character recognition system that recognizes digital text. A document is first detected using contour-based detection technique without altering the angle of the image and is segmented into lines, once the lines are segmented the words embedded in them are extracted. This segmentation is done using projection profiling method. Characters are then segmented words with vertical projection profiling from the extracted words. These characters are fed into an image recognition model for recognition. The recognition model is CNN based deep learning model. Modified VGG16 architecture is used here to extract maximum features from the images and then classify them. To train the model a dataset is created from a repository of digital character dataset. The dataset consists of images of 153 font variants. 2022 IEEE. -
An Analysis of Levenshtein Distance Using Dynamic Programming Method
An edit distance (or Levenshtein distance) amongst dual verses refers to the slightest amount of replacements, additions and omissions of signs essential to turn one name addicted to the additional is referred to as the edit distance (or Levenshtein distance) amongst dual verses. The challenge of calculating the edit distance of a consistent verbal, that is the set of verses recognised by a fixed mechanism, is addressed in this research. The Levenshtein distance is a straightforward metric for calculating the distance amongst dual words using a string approximation. After witnessing its efficiency, this approach was refined by combining certain comparable letters and minimising the biased modification between associates of the similar set. The findings displayed a considerable enhancement over the old Levenshtein distance method. 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. -
Capacity Aware Active Monitoring Load Balancing Principle for Private Cloud
Virtual machines (VMs) are the basic compute elements in cloud computing. There are load balancing principles associated with a job scheduler assigns the requests to these computing elements. Deploying an effective load balancing principle enhances better performance that ultimately achieves users satisfaction at the high level. Assigning an equal requests load appropriate to the capacity of the VMs will be a fair principle that can be the objective of any load balancing principle. Active monitoring load balancing principle assigns the requests to a server based on the pre-computed threshold limit. This paper presents a technique for assessing the capacity of the VMs based on a common attribute. This work measures each VMs processing ability as a percentage using the statistical method called Z-score. A threshold is quantified and the requests are proportioned based on this threshold value. Each server is then assigned with the proportioned requests. Suitable experiments were conducted Requests Assignment Simulator (RAS), a customized cloud simulator. The results prove that the performance of the proposed principle is comparatively better than a few load balancing principles. Areas of future extension of this work were also identified. 2021, The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. -
Data Reduction Techniques in Wireless Sensor Networks with AI
Due to their numerous uses in practically every part of life and their related problems, such as energy saving, a longer life cycle, and better resource usage, the research of wireless sensor networks is ongoing. Its extensive use successfully saves and processes a considerable volume of sensor data. Since the sensor nodes are frequently placed in challenging locations where less expensive resources are required for data collection and processing, this presents a new difficulty. One method for minimizing the quantity of sensor data is data reduction. A review of data reduction methods has been provided in this publication. The different data reduction approaches that have been put forth over the years have been examined, along with their advantages and disadvantages, ways in which they can be helpful, and whether or not using them in contexts with limited resources is worthwhile. 2022 IEEE. -
Predicting a Rise in Employee Attrition Rates Through the Utilization of People Analytics
Modern organizations have a multitude of technological tools at their disposal to augment decision-making processes, with artificial intelligence (AI) standing out as a pivotal and extensively embraced technology. Its application spans various domains, including business strategies, organizational management, and human resources. There's a growing emphasis on the significance of talent capital within companies, and the rapid evolution of AI has significantly reshaped the business landscape. The integration of AI into HR functions has notably streamlined the analysis, prediction, and diagnosis of organizational issues, enabling more informed decision-making concerning employees. This study primarily aims to explore the factors influencing employee attrition. It seeks to pinpoint the key contributors to an employee's decision to quit an organization and develop a futuristic data driven model to forecast the possibility of an employee leaving the organization. The study involves training a model using an employee turnover dataset from IBM analytics, including a total of thirty-five features and approximately one thousand and five hundred samples. Post-training, the model's performance is assessed using classical metrics. The Gaussian Nae Bayes classifier emerged as the algorithm delivering the most accurate results for the specified dataset. It notably achieved the best recall (0.54) indicating its ability to correctly identify positive observations and maintained false negative of merely 4.5%. 2023 IEEE. -
Artificial Intelligence in Disaster Management: A Survey
This paper provides a literature review of cutting-edge artificial intelligence-based methods for disaster management. Most governments are worried about disasters, which, in general, are unbelievable events. Researchers tried to deploy numerous artificial intelligence (AI)-based approaches to eliminate disaster management at different stages. Machine learning (ML) and deep learning (DL) algorithms can manage large and complex datasets emerging intrinsically in disaster management circumstances and are incredibly well suited for crucial tasks such as identifying essential features and classification. The study of existing literature in this paper is related to disaster management, and further, it collects recent development in nature-inspired algorithms (NIA) and their applications in disaster management. 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. -
Revolutionizing Arrhythmia Classification: Unleashing the Power of Machine Learning and Data Amplification for Precision Healthcare
This paper presents a comprehensive exploration of arrhythmia classification using machine learning techniques applied to electrocardiogram (ECG) signals. The study delves into the development and evaluation of diverse models, including K-Nearest Neighbors, Logistic Regression, Decision Tree Classifier, Linear and Kernelized Support Vector Machines, and Random Forest. The models undergo rigorous analysis, emphasizing precision and recall due to the categorical nature of the dependent variable. To enhance model robustness and address class imbalances, Principal Component Analysis (PCA) and Random Oversampling are employed. The results highlight the effectiveness of the Kernelized SVM with PCA, achieving a remarkable accuracy of 99.52%. Additionally, the paper discusses the positive impact of feature reduction and oversampling on model performance. The study concludes with insights into the significance of PCA and Random Oversampling in refining arrhythmia classification models, offering potential avenues for future research in healthcare analytics. 2024 IEEE. -
Review and Design of Integrated Dashboard Model for Performance Measurements
This article presents a new approach for performance measurement in organizations, integrating the analytic hierarchy process (AHP) and objective matrix (OM) with the balanced scorecard (BSC) dashboard model. This comprehensive framework prioritizes strategic objectives, establishes performance measures, and provides visual representations of progress over time. A case study illustrates the methods effectiveness, offering a holistic view of organizational performance. The article contributes significantly to performance measurement and management, providing a practical and comprehensive assessment framework. Additionally, the project focuses on creating an intuitive dashboard for Fursa Foods Ltd. Using IoT technology, it delivers real-time insights into environmental variables affecting rice processing. The dashboard allows data storage, graphical representations, and other visualizations using Python, enhancing production oversight for the company. The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024. -
CPAODV: Classifying and Assigning 3 Level Preference to the Nodes in VANET Using AODV Based CBAODV Algorithm
Vehicles communicate with nearby vehicles to share high routing and traffic information in Vehicular Ad hoc Networks (VANETs) environment. Congestion and Delay in the transmission may occur due to the density of the nodes in the network. Traffic condition depends on the vehicles in Rural and Urban environment. Increase or Decrease in vehicles speed makes significant network changes when compared to the MANET environment. Road Side Terminals (RSTs) plays a major role in bridging the connection between the sender and the receiver nodes. The traditional AODV algorithm performs better when there are shortest path and link lifetime between the nodes in VANET. Giving 3 Level Preference to the nodes as High Preference (HP), Average Preference (AP) and Less Preference (LP) gives chances to nodes that have High Preference when compared to Less Preference. CPAODV model is proposed by implementing Classifying and giving preference to the RREQ to mitigate latency to the nodes. RST sends RREQ wisely based on the early model of Route Discovery stage itself. NS2 Simulator is used to analyze the strength of the proposed algorithm using QoS metrics like Throughput, Packet Delivery Ratio and End to End delay. This proposed CPAODV method performs better when compared to traditional AODV and CBAODV algorithm. 2020, Springer Nature Switzerland AG. -
Cryptocurrencies: An Epitome of Technological Populism
From a global perspective, which holds significant cryptocurrencies, this study discusses the volatility and spillover effect between the whales cryptocurrencies. Volatility in cryptocurrency markets has always been a time-varying concept that changes over time. As opposed to the stock market, which has historically and recently, the cryptocurrency market is much more volatile. The markets have evidenced many fluctuations in the prices of cryptos. As a result, countries are transforming their policies to suit financial technologies in their economic practices. Blockchain technology allows people to obtain more benefits in a financial transaction and breaks hurdles in the financial system. The study has found no ARCH effect in BinanceCoin, BT Cash, Bitcoin, Vechain, and Zcash. It is discovered that there is an ARCH effect in the case of Ethereum, Tether, Tezos, and XRP. Whale cryptocurrencies have an ARCH effect. Daily closing prices of ten cryptocurrencies, including bitcoin, from January 1, 2019, to December 31, 2020, to determine the price volatility where the bitcoin whales hold significant cryptocurrency values. It has given significant results in case of volatility since we also found that Bitcoin's largest cryptocurrencies among the sample taken for the study have less volatility than other currencies. 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG. -
Role of Blockchain in the Healthcare Sector: Challenges, Opportunities and Its Uses in Covid-19 Pandemic
As the world grapples with the Covid-19 pandemic and major populations are getting vaccinated, increasing realisation processes healthcare industry needs to be augmented. It includes managing supply chains, healthcare records, and patient care. With a scarcity of time and resources, adaptation of blockchain technology will help mitigate the pressures on existing infrastructure. A blockchain distributed ledger helps to exchange health information securely without complex intermediation of trust with secure access. The organisations and persons in the blockchain network can verify and authorise the data, thus protecting patient identity, privacy, medical information system, and reducing transaction costs. The paper examines managing and protecting electronic medical records and personal health records data using blockchain. It also analyses issues in healthcare, blockchain implementation, and its uses in the Covid-19 pandemic. 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG. -
Optimal Shortest Path Routing over Wireless Sensor Networks Using Constrained Genetic Firefly Optimization Algorithm
In Wireless Sensor Networks (WSNs), a large number of sensor nodes are placed over a specific area in any real-life application. The sensor node is small, with limited battery life, memory, and computing capacity. Due to the limited power of the battery, WSNs must expand the system life by minimizing the energy usage. In the existing system, the methods have limitations related to optimal shortest routing path, node energy consumption, network reconfiguration, and so on. In order to overcome these issues, aConstrained Genetic FireFly Optimization Algorithm (CGFFOA) is proposed. The CGFFOA algorithm is designed to select the best shortest path routing through the selection of Cluster Head (CH) nodes based on the better energy utilization, delay, and high throughput sensor nodes. It is used to optimize the routing path based on the energy, hop count, inter and intra cluster delay, and lifetime. The simulation findings therefore conclude that, with regard to reduced energy consumption, higher throughput, and lower end-to-end delay, the proposed CGFFOA algorithm is preferable to existing methods such as Particle Swarm Optimization (PSO) and Dynamic Source Routing (DSR). 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. -
Academic workbench for streetlight powered by solar PV system using internet of everything (IoE)
Renewable energy is one of the growing trend in developing countries. Rapid development of renewable energy leads to the economic benefits and reduce environmental pollution. According to current scenario 20 to 40 percent of the power generated is consumed by streetlights. The problems faced by the current street lighting systems are when there is availability of light there is no proper utilization. Sun intensity shift is not constant all the time, it varies as the climate changes. Real time monitoring and control using intelligent algorithm avoids energy wastage during day time. ZigBee as a communication protocol current and voltage values are sent and received. Base Controller (Single Board Computer) acts as an interphase between the communication protocol and the cloud account. Remote client application is developed to control and monitor streetlight. 2018 IEEE. -
Malpractice Detection in Examination Hall using Deep Learning
Various institutions administer tests at designated examination locations, chosen third-party and approved centers, and have established standards for installing CCTV cameras and conducting frisking under the supervision of designated personnel. Some institutions are using online proctoring, which enables students to take exams from any location. In all of the aforementioned scenarios, human monitoring is conducted, and maintaining a high level of vigilance may be challenging due to administrative oversight or intentional allowance of malpractice for personal gain. The malpractice detection may be attributed to acts like as plagiarism, unauthorized sharing of papers, and non-verbal communication. The study is conducted by capturing the dataset in the classroom of Christ University. The proposed approach is based on the YOLO framework. The movies are processed in real time to identify hand rotation, paper extraction, and classify the motion. The accuracy for the Head_right class is significantly higher than that of the Head_left class. The system is implemented using the programming language Python and has the potential for future expansion to provide real-time monitoring. 2024 IEEE. -
An Advanced and Ideal Method for Tumor Detection and Classification from MRI Image Using Gamma Distribution and Support Vector Machine
As indicated by a measurable report distributed by the registry of central brain tumor at United States (CBTRUS), roughly 59,550 individuals were recently diagnosed to have essential benign and essential harmful brain tumors in 2017. Besides, in excess of 91,000 individuals, in the United States alone, were living with an essential harmful cerebrum tumor and 367,000 were living with an essential kind brain tumor. The task of detecting the position of the tumor in the body of the patient is the starting point for a medical treatment in the diagnosis process. The main aim of this study is to design a computer system, which is able to detect the tumor presence in the digital images of the brain in the patient and to accurately define its borderline. In this proposed model, gamma distribution method is used for training, testing, and for the feature extraction process, while SVM, support vector machine is used for the classification process. Most of the algorithms find it difficult to segment the tumors that were present in the edges. But with the help of gamma distribution along with the use of edge analysis, it is easier to identify those tumor areas that are present in the edges, thus making it easier for the preprocessing process. Gamma distribution also provides us with high accuracy, and it can also point the exact location of the tumor than compared to other algorithms. 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. -
Estimation of Vehicle Distance Based on Feature Points Using Monocular Vision
In this digital era safety and security have the highest precedence, the advanced driver assistance system is the latest trend and where many challenges are open for researchers. Vehicle to vehicle distance estimation is one of the most important challenges to provide the security and safety alerts for the driver. In order to achieve this, image of the front vehicle is captured using the single camera under monocular vision to estimate the vehicle distance. Then three key steps are designed to estimate the vehicle distance: extracting and locating the key features of the vehicle, characteristic triangle is drawn between those features to calculate pixel area and develop the measuring formula to calculate the distance. For efficient feature extraction and localizing of the feature position, conventional AdaBoost algorithm is utilized to find the strong features for scalable samples. Distance measurement formulation is used to derive the correlation between the pixel area and distance by considering the different parameters from the prototype of pinhole camera, camera standardization and plotting of area. Formula is developed to estimate the optimum moving distance between vehicles to vehicle. After the experimental analysis, the accuracy rate is improved and time complexity satisfies the precision. 2019 IEEE. -
Classification of Vehicle Make Based on Geometric Features and Appearance-Based Attributes Under Complex Background
Vehicle detection and recognition is an important task in the area of advanced infrastructure and movement administration. Many researchers are working on this area with different approaches to solve the problem since it has a many challenge. Every vehicle has its on own unique features for recognition. This paper focus on identifying the vehicle brand based on its geometrical features and diverse appearance-based attributes like colour, occlusion, shadow and illumination. These attributes will make the problem very challenging. In the proposed work, system will be trained with different samples of vehicles belongs to the different make. Classify those samples into different classes of models belongs to same make using Neural Network Classifier. Exploratory outcomes display promising possibilities efficiently. 2019, Springer Nature Singapore Pte Ltd. -
Classification of Vehicle Type on Indian Road Scene Based on Deep Learning
In Recent days an intelligent traffic system [ITS] is implemented on indian traffic sytem. Different applications are widely used to improvies the performance of the system. To improve the intelligence of the system deep learning can used to classify the vehicles into three different classes. The combination of Faster RCNN classifier and RPN can used to detect the objects and classify those objects into different classes. Analysis of the experimental results shows the improved accuracy and efficiency in classifying the vehicles on indian roads into different categories. 2021, Springer Nature Singapore Pte Ltd.