Real-world Applications Research Articles

Posterior Representations for Bayesian Context Trees: Sampling, Estimation and Convergence

We revisit the Bayesian Context Trees (BCT) modelling framework for discrete time series, which was recently found to be very effective in numerous tasks including model selection, estimation and prediction. A novel representation of the induced posterior distribution on model space is derived in terms of a simple branching process, and several consequences of this are explored in theory and in practice. First, it is shown that the branching process representation leads to a simple variable-dimensional Monte Carlo sampler for the joint posterior distribution on models and parameters, which can efficiently produce independent samples. This sampler is found to be more efficient than earlier MCMC samplers for the same tasks. Then, the branching process representation is used to establish the asymptotic consistency of the BCT posterior, including the derivation of an almost-sure convergence rate. Finally, an extensive study is carried out on the performance of the induced Bayesian entropy estimator. Its utility is illustrated through both simulation experiments and real-world applications, where it is found to outperform several state-of-the-art methods.

Development of an efficient vehicle-to-grid method for massive electric vehicle aggregation

The growing adoption of renewable energy and electric vehicles (EVs) has contributed to environmental sustainability; nevertheless, integration of these products into the power grid has become complex owing to their unpredictable nature and variable energy demands. A significant challenge lies in the realization of large-scale, coordinated control of EVs to serve as an alternative to traditional energy storage systems. This challenge is underscored by the complexity of optimization in large-scale cooperative control problems and the difficulty in reducing such problems to an easily manageable and practical level in real-world application. In response to these challenges, a practical mechanism for the integration of EVs into a vehicle-to-grid concept is proposed in this study. In this approach, the constraints involved in merging multiple EVs into a fictitious clustered energy storage unit, which are often neglected, are given renewed focus. An iterative multi-stage optimization method is introduced that includes an EV aggregation clustering model, multi-tier optimization model, and recursive framework. Here, marked efficacy for larger EV fleets is demonstrated for this method, providing optimal charging and discharging schedules for each vehicle while a high degree of precision is maintained. With the proposed technique, validated through numerous case studies using historical data, the global optimum solution is largely approximated, with a marginal deviation of 4 %. In addition, robustness of the model is demonstrated under varying pricing scenarios, with a computation time that increases in a linear manner, rather than exponentially, as the number of EVs increases. Meanwhile, compared with conventional methods, the technique proposed in this study has the capacity to fulfill the charging demands of all users with reduced charging expense, demonstrating the high precision and efficacy of the technique.

Pragmaticism in Cancer Clinical Trials.

Clinical trials are essential for advancing oncology treatment strategies and have contributed significantly to the decline in cancer mortality rates over the past decades. Traditional explanatory trials, focused on establishing intervention efficacy in ideal settings, often lack generalizability and may not reflect real-world patient care scenarios. Furthermore, increasing complexity in cancer clinical trial design has led to challenges such as protocol deviations, slow enrollment leading to lengthened durations of trial, and escalating costs. By contrast, pragmatic trials aim to assess intervention effectiveness in more representative patient populations under routine clinical conditions. Here, we review the principles, methodologies, challenges, and advantages of incorporating pragmatic features (PFs) into cancer clinical trials. We illustrate the application of pragmatic trial designs in oncology and discuss the QUASAR collaborative, TAPUR study, and the ongoing PRAGMATICA-LUNG trial. Although not all oncology trials may be amenable to adopting fully pragmatic designs, integration of PFs when feasible will enhance trial generalizability and real-world applicability. Project Pragmatica and similar initiatives advocate for the integration of real-world practice with clinical trials, fostering a nuanced approach to oncology research that balances efficacy and effectiveness assessments, ultimately with a goal of improving patient outcomes.

Performance analysis of different control models for smart demand–supply energy management system

Several features of innovative grid technologies can be deployed to improve the overall performance of the power system environment. This can be seen from the generation to the consumption of energy. The two-way communication of smart metering introduces the novel functionalities of the energy management system. This paper presents a practical implementation of using the intelligent metering system. It consists of implementing a nanogrid that optimally coordinates the energy from the solar panel, battery storage and utility grid to supply the end user. The developed model is validated with an optimal value of the state of charge of the distributed energy storage to maximise energy from the solar panel and battery storage while minimising the power received from the utility grid. A demand response scheme is employed to formulate the performance index of the energy management system using three optimal control models: adaptive open-loop control, adaptive closed-loop control and model predictive control schemes. The formulation of the performance index of each approach is a function of the energy flow from different resources depending on the power consumption. The three models have given different insights into the performance of the smart nanogrid, which may be used to the advantage of the grid owner or end user. Through the performance of the optimal strategies, it can be observed that energy management is ensured, and real-time monitoring of the entire system is guaranteed. The performance models facilitate the minimisation of the power from the utility, resulting in savings between 23.7% and 39.240% of the total energy demand from the end user. Besides, the system design is validated by an electrical system to form a real-world innovative nanogrid application in residential environments.

Cost optimization of electricity in energy storage system by dynamic programming

This paper presents a dynamic programming solution for the cost optimization of an electric storage system. The objective is to minimize the total cost of meeting electricity demand over a specified time interval, considering energy constraints and costs. The proposed algorithm efficiently determines the optimal energy discharge and charge strategies for the storage system, resulting in reduced overall costs. The effectiveness and efficiency of the algorithm are demonstrated through various test cases, highlighting its potential for real-world applications in energy storage systems and electric grid management. It also provides an overview of different types of electrical storage systems, review recent research on optimization techniques for energy storage, and examines recent studies on the optimization of electrical storage systems for specific applications, such as peak load shaving and grid stability. Through this comprehensive analysis, we hope to shed light on the current state of the field and identify areas for further research and improvement.

Experimental analysis for comparison of wireless transmission technologies: Wi-Fi, Bluetooth, ZigBee and LoRa for mobile multi-robot in hostile sites

This research paper conducts a thorough comparison of four prominent transmission technologies suitable for mobile robots operating in challenging environments. Emphasizing key factors such as signal strength, noise resistance, and data transfer efficiency, the study aims to identify the optimal communication solution in hostile conditions. The exploration delves into the intricacies of received signal strength indication (RSSI) and signal-to-noise ratio (SNR), revealing distinctive traits and trade-offs among the technologies. Navigating through the complexities of frequency bands, modulation types, and communication topologies, the paper examines the impact of obstacles, energy consumption dynamics, and potential real-world applications. Beyond contributing to the fields of robotics and communication, the study offers practical insights for stakeholders seeking resilient and efficient transmission methods for mobile robotic applications. Advocating for long range (LoRa) as the preferred transmission technology in hostile environments, the paper highlights its unmatched immunity to noise, stability, and minimal energy consumption. These findings provide valuable guidance for technology choices in collaborative mobile robot operations under challenging conditions. This research sets the stage for future developments in robotic communication, underscoring the crucial role of selecting the right transmission means for mission-critical applications in hostile environments.

A lightweight YOLOv5 for real-time dangerous weapons detection

Deep neural networks are currently employed to detect weapons, and although these techniques provide a high level of accuracy, it still suffers from large weight parameters and a slow inference speed. When it comes to real-world applications, such as weapon detection, these methods are often not suitable for deployment on embedded devices. Because of the huge number of parameters and poor efficiency. The most recent object detection technique, which belongs to the YOLOv5 class, is commonly used for detecting weapons. However, it faces some difficulties such as high computational parameters and an unfavorable detection rate. to solve these shortcomings. an enhanced lightweight Yolov5s approach is suggested. Which consists of a combination of YOLOv5 and GhostNet modules. To evaluate the efficacy of the suggested technique, a set of experiments was performed on the Sohas weapon dataset., which is commonly used as a reference dataset in the field. Compared to the original YOLOv5, the results indicate a slight increase in the proposed model's mean Average Precision (mAP). Furthermore, there has been a reduction of 2.7 in GFLOPs and weights, and the number of model parameters has decreased by 1.42.

Evaluating the machine learning models based on natural language processing tasks

<span lang="EN-US">In the realm of natural language processing (NLP), a diverse array of language models has emerged, catering to a wide spectrum of tasks, ranging from speaker recognition and auto-correction to sentiment analysis and stock prediction. The significance of language models in enabling the execution of these NLP tasks cannot be overstated. This study proposes an approach to enhance accuracy by leveraging a hybrid language model, combining the strengths of long short-term memory (LSTM) and gated recurrent unit (GRU). LSTM excels in preserving long-term dependencies in data, while GRU's simpler gating mechanism expedites the training process. The research endeavors to evaluate four variations of this hybrid model: LSTM, GRU, <a name="_Hlk158992630"></a>bidirectional long short-term memory (Bi-LSTM), and a combination of LSTM with GRU. These models are subjected to rigorous testing on two distinct datasets: one focused on IBM stock price prediction, and the other on Jigsaw toxic comment classification (sentiment analysis). This work represents a significant stride towards democratizing NLP capabilities, ensuring that even in resource-constrained settings, NLP models can exhibit improved performance. The anticipated implications of these findings span a wide spectrum of real-world applications and hold the potential to stimulate further research in the field of NLP. </span>

Performance evaluation of low energy adaptive clustering hierarchy-based cluster routing protocols in wireless sensor networks using a new graphical user interface

Wireless sensor network (WSN) is widely used for field data acquisition and monitoring in different domains. To make this type of network functional, efficient routing protocols must be implemented. Nevertheless, WSNs have an energy constraint due to limited batteries. Many clustered protocols are proposed to overcome it. However, the implementation of these protocols would be difficult to understand without a simulation tool, as some problems may arise during their development. Testing real-world applications requires a lot of effort and cost because they often use many nodes in large networks. Therefore, the simulation tool is the most relative way to evaluate these protocols. This paper presents graphical-based cluster protocols simulation interface for WSN(GCPS-WSN), a new interface to simulate some clustered protocols in WSNs. GCPS-WSN allows the user to evaluate the performance of certain low energy adaptive clustering hierarchy (LEACH) enhanced protocols to choose the most appropriate one for his system. The user can simulate protocols without any knowledge of software programming.

Advanced method for Image detection using OpenCV

Abstract: In computer vision, object detection is an essential job with many real-world applications, such as robots, autonomous cars, and surveillance. Because of its great accuracy and speed, the YOLO (You Only Look Once) method is a well-liked realtime object recognition technique that has attracted a lot of attention. This technique is perfect for time-sensitive applications since it examines the full image at once and predicts bounding boxes and class probabilities for recognized items. YOLO has undergone many iterations, with YOLO v5 being the most recent and sophisticated version. It utilizes anchor boxes and a feature pyramid network (FPN) to increase the accuracy of object recognition. Our goal in this research is to apply YOLO v5 to realtime picture and object identification applications stage. The model will be trained using an appropriate dataset, and its performance will be assessed against a range of benchmarks and advanced object detection methods. The project's output will offer a reliable and effective real-time object detection system that will facilitate prompt decision-making in the identification of item types and their corresponding placements. It has useful uses in robotics, autonomous driving, and surveillance.

The Impact of Social Media on Consumer Behavior and Purchasing Decisions in the Fashion Industry

Abstract: The present study aims to examine the noteworthy impact of social media platforms on customer behavior and purchase decisions in the ever-changing fashion business. Fashion marketers now need to grasp how social media usage affects consumer behavior due to its exponential expansion worldwide, especially among fashion-conscious populations. This study investigates the complex relationship between social media involvement and several aspects of consumer behavior, such as brand awareness, brand loyalty, buy intentions, and actual purchase behavior. It does this by conducting a thorough review of the existing literature and doing empirical analysis. Additionally, the study explores the ways in which brand advertising, influencer endorsements, and user-generated material on social media influence customer perceptions and preferences in the fashion industry. This study integrates theoretical understandings with real-world applications is to offer insightful information to fashion firms and marketers who want to improve their social media tactics and gain a competitive edge in a sector that is becoming more and more digital

Privacy-Centric AI: Navigating the Landscape with Federated Learning

Abstract: In the era of big data and privacy concerns, federated learning has emerged as a promising approach to training machine learning models while preserving data privacy. This paper explores the principles and applications of federated learning, highlighting its potential to revolutionize privacy-centric AI. We discuss the methodology, significance, and challenges of federated learning, providing insights into its future directions. By leveraging decentralized data and aggregating model updates, federated learning enables the development of powerful AI models without compromising individual privacy. We present real-world applications and cite relevant studies to demonstrate the transformative impact of federated learning across various domains.

Automating Helmet Usage Detection: A YOLOv8 Based Framework

Abstract: The biggest threat to Powered two-wheeler(PTW) riders is a head injury. With the humongous amount of daily commuters on such PTWs, the risk only increases substantially. In such cases, helmets reduce the risk of head injuries in accidents but not every citizen is keen on their use. In India, where the use of motorcycles is widespread, ensuring helmetuse remains a challenge. This study explores the effectiveness of YOLOv8(You Only Look Once) a state-of-the-art single-stage object detection algorithm for helmet detection on Indian roads [14]. The CNN(Convolutional Neural Networks) based technologyimplements many other efficient, reliable and quick algorithmsto train, validate and predict the object’s detection, segmentationand classification task. We leverage the use of a publicly available Indian helmet dataset from Kaggle, containing 942 images with annotations. This an image dataset of real-life powered two-wheeler riders. The dataset has images and video frames for training on 5 different objects/classes along with annotations that specify thoseobjects. This multi-class approach offers valuable insight intohelmet usage patterns paves shows us the way for real-world applications for automated traffic monitoring. Real-time helmet and number plate detection can become a game changer in trafficmonitoring and can strengthen basic safety laws

Application of AI in the Field of Education: Dynamic Course Generation

Abstract: This research article examines the evolving relationship between Artificial Intelligence (AI) and Education, exploring its impact and potential. It covers adaptive learning systems, intelligent tutoring platforms, and emerging technologies, analyzing real-world applications and addressing ethical and implementation challenges. By synthesizing research findings, it contributes to understanding AI's role in transforming learning paradigms

Review of Challenges and Innovations in Occluded Facial Recognition in Disguise and Crowd

Abstract: Facial recognition technology is critical in modern security and identification systems, yet it faces significant challenges when dealing with occluded faces, whether due to disguises or in crowds. This review explores the complexities of occluded facial recognition, highlighting the obstacles presented by partial or complete obstruction of facial features by various elements like clothing, accessories, or other people in crowded spaces. These occlusions, intentional or otherwise, complicate the identification process, rendering traditional recognition systems less effective. Recent advancements in machine learning, particularly through Convolutional Neural Networks (CNNs) and Generative Adversarial Networks (GANs), have ushered in innovative solutions that enhance the system's ability to recognize faces even when significant portions are obscured. These developments signify a move towards more adaptable and precise occluded facial recognition technologies, essential for real-world applications where occlusions are unpredictable. The paper discusses ongoing challenges and innovations considerations, various occluded system performance metrics and conclude with way ahead for Researchers and Practioners

Medical Symptom Text Classification through NLP in Healthcare

Abstract: Medical symptom text classification through Natural Language Processing (NLP) is a rapidly evolving field that aims to leverage computational techniques to analyse and interpret vast amounts of textual data generated in healthcare settings. This paper provides a comprehensive survey of current methodologies, applications, challenges, and future directions in this domain. We begin by discussing the importance of symptom classification for improving patient outcomes, supporting clinical decisionmaking, and enhancing disease surveillance. We then review traditional machine learning approaches and advanced deep learning models, highlighting their respective strengths and limitations. Key pre-processing techniques crucial for handling medical jargon and ensuring data privacy are also examined. The paper further explores real-world applications, including clinical decision support systems, disease outbreak detection, and patient monitoring. Despite significant advancements, challenges such as data quality, model interpretability, and regulatory compliance remain. Finally, we identify emerging trends and potential future developments that could drive further innovation in NLP for healthcare. This survey aims to provide a valuable resource for researchers and practitioners seeking to understand and contribute to the field of medical symptom text classification.

Exploring Sentiment Analysis across Text, Audio, and Video: A Comprehensive Approach and Future Directions

Abstract: This study presents a comprehensive exploration of sentiment analysis techniques across text, audio, and video modalities. Leveraging natural language processing (NLP), speech recognition, and computer vision algorithms, the research demonstrates the versatility and adaptability of sentiment analysis across diverse data sources. The necessity of such an approach lies in its ability to provide deeper insights into user emotions and opinions expressed in various mediums, including written text, spoken language, and visual content. Moreover, the study highlights the importance of sentiment analysis in understanding customer feedback, market trends, social media sentiments, and sentiment-aware recommendation systems. Future directions include advancing algorithmic accuracy and efficiency, integrating multimodal fusion techniques, and exploring applications in diverse domains, thereby paving the way for enhanced sentiment analysis capabilities and broader realworld applications

LipNet: End-to-End Lipreading

Lipreading is the task of decoding text from the movement of a speaker’s mouth. This research presents the development of an advanced end-to-end lipreading system. Leveraging deep learning architectures and multimodal fusion techniques, the proposed system interprets spoken language solely from visual cues, such as lip movements. Through meticulous data collection, annotation, preprocessing, model development, and evaluation, diverse datasets encompassing various speakers, accents, languages, and environmental conditions are curated to ensure robustness and generalization. Conventional methods divided the task into two phases: prediction and designing or learning visual characteristics. Most deep lipreading methods are trainable from end to end. In the past, lipreading has been tackled using tedious and sometimes unsatisfactory techniques that break down speech into smaller units like phonemes or visemes. But these methods often fail when faced with real-world problems, such contextual factors, accents, and differences in speech patterns. Nevertheless, current research on end-to-end trained models only carries out word classification; sentence-level sequence prediction is not included. LipNet is an end-to-end trained model that uses spatiotemporal convolutions, a recurrent network, and the connectionist temporal classification loss to translate a variable-length sequence of video frames to text. LipNet breaks from this traditional paradigm by using an all-encompassing, end-to-end approach supported by deep learning algorithms, Convolutional neural networks (CNNs) and recurrent neural networks (RNNs), which are skilled at processing sequential data and extracting high-level representations, are fundamental to LipNet's architecture.LipNet achieves 95.2% accuracy in sentence-level on the GRID corpus, overlapped speaker split task, outperforming experienced human lipreaders and the previous 86.4% word-level state-of-the-art accuracy.The results underscore the transformative potential of the lipreading system in real-world applications, particularly in domains such as assistive technology and human-computer interaction, where it can significantly improve communication accessibility and inclusivity for individuals with hearing impairments.

Adaptation strategies of paddy farmers to build resilience to climate change in kahama district, Tanzania

The study investigates the adaptation strategies used by the paddy farmers in Kahama District to increase their adaptability to climate change. The study focuses on four main objectives, including changed farming practices, changes in cropping practices, strategies used to adapt to climate change, and desired interventions to improve agricultural production and resilience to climate change. A mixed-methods strategy is used in the study to collect both quantitative and qualitative data. According to the data, paddy farmers in Kahama District have used a number of climate change adaptation strategies. These adaptations include modifications to irrigation techniques, timing adjustments for planting, seed variety changes, and fertilizer treatment adjustments. The study also reveals a change in crop systems, with farmers choosing drought-tolerant cultivars, intercropping, and crop diversification; increased water management practices, increased soil conservation practices, and informed decision-making through the use of climate information services are some of the strategies used by paddy farmers to adapt to climate change. The study also identifies the interventions that farmers want to see implemented, such as enhanced irrigation systems, workshops on climate-resilient farming techniques, access to resilient seed varieties, and dependable weather forecasting and climate information services. These findings show the paddy farmers in the Kahama District's proactive response to the problems posed by climate change and help to better understand the adaptation measures they have used. The study highlights the importance of assisting farmers in advancing their current strategies and knowledge while also attending to their unique requirements. Implementing desired actions, such as expanding climate information services, supplying training and access to robust seed varieties, and enhancing irrigation infrastructure, can significantly increase agricultural productivity and the region's resistance to climate change. For politicians, extension services, and development organizations looking to promote climate change adaptation and sustainable agriculture, these research findings have real-world applications. Paddy farmers' ability to adapt, agricultural output, and livelihood security in the face of climate change problems can all be improved by matching interventions and policies with their recognized needs and goals. Such activities can support attempts to achieve food security and sustainable development in the Kahama District and other areas, as well as the agricultural sector's overall resilience.

Classification of gram-positive and gram-negative bacteria using Few-shot learning algorithm

Our research introduces a novel approach to classify bacteria as Gram-positive or Gram-negative using few-shot learning. We employ deep neural networks, specifically Prototypical Networks, to learn distinctive features from bacterial images, enabling accurate classification even with limited data. Experimental results on diverse datasets demonstrate the model's effectiveness and potential for real-world applications in microbiology and healthcare. We also address interpretability, ethics, and data privacy, making it a valuable tool for bacterial classification and diagnostics.