International Journal on Recent and Innovation Trends in Computing and Communication
Not a member yet
8613 research outputs found
Sort by
A Novel Hybrid Deep Learning System for Cardiovascular Detection and Salient Feature Extraction from ECG Data
Get PDFThe Cardiovascular system is responsible for the circulation of blood throughout the body. Abnormalities in the cardiovascular system can lead to various diseases such as arrhythmia, heart failure, and myocardial infarction. The Electrocardiogram (ECG) is unique and commonly used diagnostic tools for detecting cardiovascular diseases. The conventional methods of ECG analysis require expert interpretation and are time-consuming. Automated ECG analysis systems cantered on machine learning and also deep learning techniques have been proposed to overcome the limitations of conventional methods. In this research, we propose a hybrid deep learning-based cardiovascular detection system that can accurately detect various cardiovascular diseases by extracting salient features from ECG data. The suggested approach combines feature extraction using a convolutional neural network with wavelet transform and principal component analysis. The fused signals obtained from the previous steps are optimized using Sequential Minimal Optimization (SMO) algorithm to improve classification accuracy. Therefore, the development of a reliable and automated ECG analysis system is highly desirable testing on a publicly available ECG dataset from MIT-BIH Arrythmia
Legal Implications of Artificial Intelligence in Criminal Justice Systems
Get PDFThe rapid advancement of Artificial Intelligence (AI) is transforming various sectors, including the criminal justice system, where it is increasingly utilized for tasks such as law enforcement, sentencing, and predictive policing. AI holds the potential to enhance efficiency, accuracy, and decision-making within these areas. However, its deployment also brings forth critical ethical and legal challenges. Key concerns include accountability for AI-driven decisions, algorithmic bias, transparency in AI processes, and the diminishing role of human oversight. This research paper critically examines these challenges, explores the existing legal frameworks governing AI in criminal justice, and proposes strategies to mitigate associated risks. By addressing the ethical and legal implications of AI integration in criminal justice, this paper seeks to contribute to the evolving discourse on balancing innovation with the preservation of justice and fairness
Data-Efficient Vision: Exploring Few-Shot Learning Techniques
Get PDFFew-shot learning (FSL) enables models to generalize from limited training datasets, transforming the fields of machine learning and computer vision. We examine few-shot learning and computer vision. Few-shot learning is necessary when data is limited or expensive, and models underperform due to insufficient training samples. This abstract addresses meta-learning, metric learning, and transfer learning. Image classification, anomaly identification, and few-shot learning object detection.
Employing metrics, few-shot learning categorizes and distinguishes analogous cases. Analyze and categorize exemplary instances. Triplet loss Siamese networks enhance facial and signature authentication. Understanding distance and forecasting similarity may assist models in generalizing from limited cases.
Few-shot learning prioritizes meta-learning through the process of learning to learn. This approach enables models to rapidly adjust to new tasks with less data by leveraging prior task experience. MAML and Prototypical Networks instruct models on several tasks using few samples. MAML facilitates rapid tuning of model parameters with minimal training, addressing emerging challenges.
Transfer Learning use a substantial dataset and a pre-trained model to improve task performance with minimal data. Few-shot transfer learning utilizes a large dataset model to train on a limited number of samples. This method utilizes acquired representations for new tasks and enhances model generalization through domain adaptation and fine-tuning.
Utilitarian Numerous computer vision applications employ few-shot learning. Few-shot learning enables models to acquire new knowledge with little annotated samples, addressing the challenges of expensive or impractical data collection. Models categorize images with minimal training data through practical few-shot learning. Few-shot learning can identify anomalies in brief atypical data.
Case examples illustrate the efficacy of few-shot learning. In instances where the detection of several samples is challenging, medical image analysis utilizes few-shot learning. Meta-learning may assist in the identification of rare diseases with little medical data and photos.
The paper addresses issues of model overfitting, scalability, and generalization in few-shot learning problems. Models trained on limited examples may overfit, excelling on training data while underperforming on unfamiliar data. Effective few-shot learning systems require innovative concepts and ongoing research.
Future research in few-shot learning emphasizes generalization, scalability, and interpretability. Meta-learning, domain adaptability, and distinctive metric learning may facilitate few-shot learning. These concerns want additional examination to enhance few-shot learning and the profession
Increasing the Sale of Long Tail Items on E-Commerce Websites
Get PDFThe advent of e-commerce has transformed the retail industry, offering a platform for retailers to reach a vast audience, enabling consumers to shop from anywhere in the world at any time. However, most e-commerce websites are dominated by sales of popular or mainstream items, referred to as the head of the distribution. The long-tail items, which make up the tail of the distribution, are often overlooked, leading to reduced sales and low revenue for online retailers. This research paper aims to explore strategies that online retailers can use to increase the sale of long-tail items on their websites. It also includes a review of existing literature, a qualitative analysis of consumer behavior, and a quantitative analysis of sales data. The findings indicate that online retailers can increase the sale of long-tail items by optimizing their website design, improving search functionality, using data-driven pricing strategies, and implementing targeted marketing campaigns. These strategies have the potential to improve the visibility of long-tail items, increase consumer engagement, and boost revenue for online retailers. A study has also been conducted on the Retail Rocket e-commerce dataset from Kaggle which involved a meticulous examination of various aspects of user interactions within the online retail platform. The dataset provided a rich source of information, including events such as page views, add-to-cart actions, and completed purchases. The analysis aimed to uncover specific patterns and trends related to these events, shedding light on how users engage with both popular and long-tail items
“Shopper Mindfulness and Acuity Near Green Marketing in India”
Get PDFGreen marketing is the advertising of crops that are supposed to be ecologically benign. It joins a comprehensive series of doings including creation alteration, vagaries to the making development, justifiable wrapper as well as changing advertisement. Thus, green selling refers to full selling concept that making, eating and discarding of foods and facilities chance in a routine that is less injurious to the location with emergent alertness about the implication of global warming, non-biodegradable solid surplus, detrimental power of toxins, etc. both sellers and users are pretty increasingly subtle to the need for green products and services.
While the shift of ‘green’ may seem to be luxurious in short run, it will prove to be essential and useful in the long run. This study is essential to classify whether the customers in India are aware of green marketing and green foodstuffs. It classifies the general ecological beliefs of customers. It inspects the factors moving procurement performance of customers for green crops. It focuses the reasons which make the customers keen to pay for green products. For this purpose, primary data were collected from 150 sample respondents using a well – structured survey. Boards were used to analyse the data and use Desks, humble proportions to take it
Investigation of Enhanced Performance in Flexible Solar Cells Using Passive Cooling Technique
Get PDFThe lack of flexibility and enormous weight in conventional photovoltaic (PV) modules limits their applications. The advantages of flexibility and lightweight have made flexible solar cells popular in various applications. However, flexible PVs have an efficiency degradation due to an increase in module temperature through incoming solar infrared radiation. Both the power output and the electrical efficiency of the PV module depend linearly on the operating temperature. For every degree increase in the PV temperature, the efficiency decreases by 0.45-0.65%. Here, the novel concept of applying a nanomaterial-based heat-resistant coating for the passive cooling of flexible solar cells was experimentally investigated. A heat-resistant coating generally keeps buildings cooled by filtering UV and infrared rays and transmitting visible rays. This approach works by controlling the incoming solar radiation, thereby decreasing the overall temperature of flexible solar cells passively without adding much weight. Here, a transparent flexible polyacrylic sheet 0.25 mm in thickness was used, and two coats of silver nanomaterial-based coating were applied. The sheet was placed over a flexible solar photovoltaic module with a power rating of 6 watts. The temperature of the flexible solar photovoltaic module was recorded at different time intervals for August, September, and October using temperature sensors, taking note of factors such as wind speed and solar irradiation. These readings were compared with those taken from the solar panel without any coating. A temperature reduction of 6-7°C and an improved solar power efficiency of 2.5-4 % were observed for cooled flexible solar panels
Facial Data Classification Through Enhanced Local Binary Patterns (LBP) and Dynamic Range Local Binary Patterns (DRLBP) Algorithms
Get PDFA significant amount of reliance is placed on facial data classification in contemporary computer vision and pattern recognition. This research presents a novel method that makes use of the algorithms of Dynamic Range Local Binary Patterns (DRLBP) and Enhanced Local Binary Patterns for the purpose of face data classification that is both effective and precise (LBP). The classic LBP methodology is expanded upon by the Enhanced LBP method, which incorporates adaptive thresholding techniques and spatial histogram characteristics. This makes it possible to conduct a more thorough investigation of the texture and to be resilient in a variety of lighting conditions. By continuously modifying the local binary pattern range, the DRLBP algorithm improves upon this in order to better accommodate nuanced facial features and expressions. This is done in order to better accommodate facial expressions. In terms of accuracy, speed, and adaptability, our proposed system beats state-of-the-art alternatives, as demonstrated by extensive trials conducted on commonly used facial datasets. According to the findings of our investigation, it would appear that human-computer interaction (HCI), digital forensics, and security systems could all stand to gain a great deal from a solution that combines Enhanced LBP and DRLBP algorithms for the classification of face data
A Study on Effective Mathematics Education Through Various Modelling Strategies
Get PDFIn India, as well as many other countries, mathematical modelling and its application in mathematics education are gaining popularity. The growing body of literature on the subject uncovers a variety of techniques to mathematical modelling as well as similar theories, and also perspectives on the use of mathematical modelling in mathematical learning, in definitions of meanings of designs and modelling, conceptual backdrops of modelling, and the essence of questions used in teaching modelling. This journal is about mathematical modelling, or, to put it another way, implementations and modelling. Beginning with Henry Pollak's research lecture, this has been a significant theme in math education over several decades. The term "applications and modelling" refers to both the products and the processes involved in the interaction of the physical world and mathematics. The purpose of this paper is to summarise some critical parts, particularly those related to implementation teaching and modelling
Fruit Grade Classification and Disease Detection using Deep Learning Techniques
Get PDFEnsuring optimal food quality and agricultural productivity hinges on effective fruit quality assessment and disease detection. Introducing a comprehensive strategy employing deep learning techniques to address critical aspects of fruit quality assessment and disease detection in agriculture. The methodology is structured into two distinct phases, each designed to optimize the accuracy and efficiency of the overall system. In the initial phase, image acquisition, preprocessing, and precise Region of Interest (ROI) detection using the Expectation-Maximization (EM) method lay the foundation for fruit classification with the AlexNet architecture. Rigorous training and testing procedures ensure the model's efficacy. The subsequent phase extends the initial process, with a heightened focus on feature extraction facilitated by DenseNet201. Thorough performance analysis, incorporating multiple metrics, assesses the accuracy and effectiveness of the system. This framework aspires to establish a robust solution for automated fruit grading and disease detection. By harnessing the capabilities of deep learning models, the goal is to accurately classify fruits and identify potential diseases, contributing significantly to agricultural practices and food quality management. The anticipated outcomes aim to set the groundwork for future advancements in the agricultural sector, providing a technological solution that enhances efficiency in fruit quality assessment and disease detection, ultimately benefiting food quality and crop yield
Conceptualizing Sustainable Smart Country: Understanding Its Dependency on Smart Security Structure.
Get PDFThis paper explores the concept of a sustainable smart country and its dependence on smart security structures. It aims to understand the relationship between sustainable development and smart security and how the latter can contribute to the former. The paper defines a sustainable smart country and its key features, examines the role of smart security in achieving sustainable development goals, and analyzes the challenges and opportunities associated with implementing a smart security structure in a sustainable smart country. The research methodology involves a comprehensive literature review of relevant academic and policy sources. The findings will contribute to the ongoing debate on the role of smart security in sustainable development and provide insights for policymakers, researchers, and practitioners. It summarizes the study's purpose, basic design, major findings, interpretations, and conclusions. The research methodology is also highlighted, and the study's potential contribution to the ongoing debate on smart security in sustainable development is highlighted