Browse Items (2150 total)
Sort by:
-
Leveraging Model Distillation as a Defense Against Adversarial Attacks Based on Deep Learning
Adversarial attacks on deep learning models threaten machine learning system security and reliability. The above attacks use modest data alterations to produce erroneous model results while being undetected by humans. This work suggests model distillation to prevent adversarial perturbations. The student model is taught to emulate the teacher model in model distillation. This is done using teacher model soft outputs. Our idea is that this strategy organically strengthens the student model against adversarial assaults by keeping the teacher model's essential knowledge and generalization capabilities while reducing weaknesses. Distilled models are more resilient to adversarial assaults than non-distilled models, according to experiments. These models also perform similarly on undamaged, uncorrupted data. The results show that model distillation may be a powerful defense against machine learning adversaries. This method protects model resilience and performance. 2023 IEEE. -
SARIMA Techniques for Predictive Resource Provisioning in Cloud Environments
Seasonal Autoregressive Integrated Moving Average (SARIMA) models for dynamic cloud resource provisioning are introduced and evaluated in this work. Various cloud-based apps provided historical data to train and evaluate SARIMA models. The SARIMA(1,1,1)(0,1,1)12 model has an MAE of 0.056 and an RMSE of 0.082, indicating excellent prediction ability. This model projected resource needs better than other SARIMA settings. Sample prediction vs. real study showed close congruence between projected and observed resource consumption. MAE improved with hyperparameter adjustment, according to sensitivity analysis. Moreover, SARIMA-based resource allocation improved CPU usage by 12.5%, RAM utilization by 20%, and storage utilization by 21.4%. These data demonstrate SARIMA's ability to forecast cloud resource needs. SARIMA-based resource management might change dynamic cloud resource management systems due to cost reductions and resource usage efficiency. This research helps industry practitioners improve cloud-based service performance and cost. 2023 IEEE. -
Artificial Intelligence, Smart Contracts, and the Groundbreaking Potential of Blockchain technology: Unlock the Next Generation of Innovation
The blockchain technology consists of blocks and is a decentralized network of nodes (miners). Each block is made up of three parts: the data, the hash, and the hash from the previous block. After data has been stored, it is extremely difficult to temper the data. Transactions are verified by miners, who are compensated with a commission for their labor. Readers will gain a comprehensive understanding of blockchain technology from this review article, including how it may be used in a variety of industries including supply chains, healthcare, and banking. Most individuals were already familiar with Bitcoin as one of the well-known blockchain applications. In this section, we'll discuss a few of the countless research publications on the cutting-edge applications of this technology. We'll talk about the challenges that come with actually using these applications as well. Blockchain is an industry that is growing thanks to its more recent applications in a number of fields, such as hospital administration, cryptocurrency use, and other places. Only the manner that blockchain works and runs makes it possible for these applications. 2023 IEEE. -
Remote Diabetic Retinopathy Screening with IoT and Machine Learning on Edge Devices
This study presents a novel method of screening for diabetic retinopathy using edge devices the Internet of Things and machine learning. The developed remote screening system ensures broad accessibility as well as affordability by overcoming geographical barriers. While edge computing maximizes real-time analysis, the integration of sophisticated machine learning algorithms improves diagnostic accuracy. The investigation of socio-technical subtleties is guided by the interpretivist philosophy. The outcomes show a strong architecture, effective models, as well as revolutionary effects on accessibility. A critical assessment finds the good points and continuous improvements. Suggestions place a strong emphasis on scaling issues and the ongoing improvement of machine learning models. In order to secure data management and keep up with changing healthcare needs, future research suggests combining blockchain technology with sophisticated imaging modalities. This study advances early detection, enhances accessibility to healthcare, and advances remote screening technologies. 2023 IEEE. -
Hyperledger Fabric as a Secure Blockchain Solution for Healthcare 4.0 Framework
The healthcare sector deals with extremely sensitive information that must be administered in a safe and confidential way. The objective of the proposed framework is to utilize Blockchain Technology (BT) for tracking medical prescriptions and the implementation is carried out using the Hyperledger Fabric platform, an enterprise-grade open-source distributed ledger technology platform designed for Bigdata applications. Multiple entities, including patients, e-pharmacies, pharmacies, doctors and hospitals can establish connections by introducing several nodes in the Fabric chain. A web-centered application is provided for doctors, connecting them with participating pharmacies, hospitals and e-pharmacies through which, they can share patient prescription. Pharmacies and e-pharmacies have access to this data and can notify patients about the availability of prescribed medicines. Additionally, reminders for refills, such as heart medication, can be sent for patients requiring long-term medication. Patients can also try with nearby pharmacies and the availability of their prescribed medicines. The inclusion of a wallet feature in the application enables patients to use mobile tokens for making purchases. Patient data is treated with the utmost confidentiality, kept private, and accessed only upon request and with the consent of the concerned parties. This privacy is ensured through the use of zero-knowledge proof. Patients retain access to their complete medical history, facilitating interactions with doctors without the need for repetitive information sharing. 2023 IEEE. -
A Machine Learning Model for Augmenting the Media Accessibility for the Disabled People
In an era characterized by the proliferation of digital media, the need to efficiently use multimedia content has become paramount. This article discusses an innovative technique called 'Fast Captioning (FC)' to improve media accessibility, especially for people with disabilities and others with time restrictions. Modern Machine Learning (ML) algorithms are incorporated into the framework, which speeds up video consumption while maintaining content coherence. The procedure includes extracting complex features like Word2Vec embeddings, part-of-speech tags, named entities, and syntactic relationships. Using annotated data, a ML model is trained to forecast semantic similarity scores between words and frames. The predicted scores seamlessly integrate into equations that calculate similarity, thus enhancing content comprehension. Through this all-encompassing approach, the article offers a comprehensive solution that balances the requirements of contemporary media with the accessibility requirements of people with disabilities, producing a more inclusive digital environment. Machine Learning-based Media Augmentation (ML-MA) has achieved the highest accuracy of 96%, and the captioning is accurate. 2023 IEEE. -
Smart Product Packing and IoT Marketing: Enhancing Customer Interaction
The convergence of smart product packaging and IoT marketing has transformed commerce. This study examines the fundamental ramifications of convergence and its potential to improve customer engagement. Our research shows the transformational potential of these technologies via quantitative and qualitative analyses.Smart packaging outperforms non-smart items, giving firms an advantage, according to quantitative data. Regression and correlation analysis confirm IoT data-customer interaction. Our study also emphasizes ethical data acquisition, which supports data privacy and consumer protection.Consumers may expect personalized experiences, transparency, and real-time feedback from this technology transformation. Smart product packaging and IoT marketing enable readers to make educated decisions and influence product development to meet changing consumer expectations.This research allows academics to study the ideas and models that affect consumer engagement. Data privacy and consumer protection may inform IoT marketing and smart device packaging policy.Our research guides organizations and customers towards better customer interactions, data-driven decision-making, and ethical data practices in this changing age. The future promises revolutionary customer contact. 2023 IEEE. -
Interpreting Scope of Predictive Analytics in Advanced Driving Assistant System
Distracted driving, caused by various factors such as human emotions or reading distracting messages on the roadside, has become a leading cause of traffic accidents today. Ensuring the safety of both individuals and vehicles while minimizing maintenance costs poses a significant challenge for the automotive industry. Fortunately, recent advancements in machine learning offer a potential solution. One promising method is the further development of Advanced Driver Assistance Systems (ADAS), for which machine learning serves as an ideal solution. The proposed model develops an advanced predictive learning enabled driving assistance system with prediction capabilities like traffic light behavior and parking availability detection. The model gave an optimum accuracy of 98.2% with 50 epochs count and the validation loss retains a constant value of 0.3 over epochs. 2023 IEEE. -
Selection of cobot for human-robot collaboration for robotic assembly task with Best Worst MCDM techniques
Since the first industrial robot was produced at the beginning of the 1960s, robotic technology has completely changed the sector. Industrial robots are made for various tasks, including welding, painting, assembling, disassembling, picking and placing printed circuit boards, palletizing, packing and labeling, and product testing. Finding flexible solutions that allow production lines to be swiftly re-planned, adjusted, and structured for new or significantly modified product development remains a significant unresolved problem. Today's Industrial robots are still mostly pre-programmed to do certain jobs; they cannot recognize mistakes in their work or communicate well with both a complicated environment and a human worker. Full robot autonomy, including organic interaction, learning from and with humans, and safe and adaptable performance for difficult tasks in unstructured contexts, will remain a pipe dream for the foreseeable future. Humans and robots will work together in collaborative settings such as homes, offices, and factory setups to execute various object manipulation activities. So, it is necessary to study the collaborative robots (cobots) that will play a key role in human-robot collaborations. Multiple competing variables must be considered in a thorough selection process to assess how well industrial cobots will work on an industrial working floor. To select a collaborative robot for the human-robot collaborative application, a straightforward multi-criteria decision-making (MCDM) methodology is based on the best-worst method (BWM). The ranking derived using the BWM method is displayed. The outcomes demonstrated the value of MCDM techniques for cobot selection. 2023 IEEE. -
Enhancing Kubernetes Auto-Scaling: Leveraging Metrics for Improved Workload Performance
Kubernetes is an open-source production-grade container orchestration platform, that can enable high availability and scalability for various types of workloads. Maximizing the performance and reducing the cost are two major challenges modern applications encounter. To achieve this, resource management and proactively deploying resources to meet specific application requirements becomes utmost important. Adopting Kubernetes auto-scaler to fit one's needs are important to maximize the performance. This study aims to perform a comprehensive analysis of Kubernetes auto-scaling policies. This paper also lists out the various parameters considered for auto-scaling, and prediction methods used to efficiently handle resource requirements of applications. The research findings reveal a scarcity in the existing work regarding the variety of workload based auto-scaling and custom metrics. This paper provides a concise overview of a forthcoming research endeavor that explores the utilization of custom metrics in conjunction with auto-scaling. 2023 IEEE. -
Enhancing Security and Resource Optimization in IoT Applications with Blockchain Inclusion
The rapid proliferation of Internet of Things (IoT) devices has ushered in a new era of connectivity and data-driven applications. However, optimizing the allocation of resources within IoT networks is a pressing challenge. This research explores a novel approach to resource optimization, combining blockchain technology with enhanced security measures, while addressing the critical concerns of time and energy consumption. In this study, we propose a resource allocation framework that leverages the transparency and immutability of blockchain to enhance data integrity and security in IoT applications. The blockchain-based method is utilized to identify the malicious users in the IoT applications. The proposed method is implemented in MATLAB and performance is evaluated by performance metrics such as the probability of detection, false alarm probability, average network throughput, and energy efficiency. The proposed method is compared by existing methods such as Friend or Foe and Tidal Trust Algorithm. To further optimize this process, we introduce a Hybrid Artificial Bee Colony-Whale Optimization Algorithm (ABC-WOA), a powerful optimization technique designed to minimize time delays and energy consumption in IoT environments. Our findings demonstrate the effectiveness of the proposed approach in achieving resource efficiency, reducing time and conserving energy within IoT networks. 2023 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. -
Intelligent Safety Life Jacket Using Wsn Technology
The body loses heat in hypothermia because it cannot maintain its internal temperature owing to a freezing environment. As a result, the body temperature will decrease rapidly. The person will lose consciousness or faint when the body temperature falls below 35C. This study targets detecting climbers' hypothermia and transmitting their health status to the climber's group. It is difficult for mountain climbers to check their health and hypothermia symptoms for themselves and their climbing companions. To address this issue, we created a life jacket with an integrated hardware kit with a Peltier, temperature, and pulse sensor. LoRa Network is used to communicate with the climber's group. Alert messages are delivered to mountaineers via the Android app and suitable protocols, which helps save the climbers if any discrepancies occur. 2023 IEEE. -
Information Extraction Using Data Mining Techniques For Big Data Processing in Digital Marketing Platforms
In the dynamic landscape of digital marketing, harnessing the potential of big data has become paramount for informed decision-making. This study explores the integration of data mining techniques within big data processing frameworks to extract valuable information in digital marketing platforms. With the exponential growth of data generated through online interactions, social media, and e-commerce, traditional methods fall short of uncovering meaningful insights. This research focuses on leveraging advanced data mining algorithms to sift through vast datasets, identifying patterns, trends, and user behaviours. The proposed approach aims to enhance marketing strategies by extracting actionable intelligence from diverse data sources. Techniques such as association rule mining, clustering, and sentiment analysis will be employed to unveil hidden correlations, segment target audiences, and gauge consumer sentiment. The scalability of big data frameworks ensures efficient processing of massive datasets, allowing marketers to make real-time, data-driven decisions. Additionally, the study explores the challenges and opportunities associated with implementing data mining in big data environments for digital marketing. This research contributes to the evolving field of digital marketing by providing a comprehensive framework for extracting, processing, and utilizing information from big data. The findings promise to empower marketers with a deeper understanding of consumer behaviour, enabling the development of more personalized and effective marketing strategies in the ever-evolving digital ecosystem. 2023 IEEE. -
Empirical study on The Role of Machine Learning in Stress Assessment among Adolescents
Stress is a psychological condition that people who are experiencing difficulties in their social and environmental well-being face, and it can cause several health problems. Young individuals experience major changes during this crucial time, and they are expected to succeed in society. It's critical for people to master appropriate stress management techniques to ensure a smooth transition into adulthood. The transition to new settings, lifestyles, and interactions with a variety of people, things, and events occurs during adolescence. In this study, a dataset was utilized to classify 520 Indian individuals' stress levels into three categories: normal, moderate, and severe. Support Vector Machines, KNN, Decision Trees, Naive Bayes and CNN were among the different classification techniques that were taken into consideration. The CNN Algorithm was found to be the most reliable method for categorizing diseases linked to mental stress. The study's main goal is to create a classification model that can correctly classify a variety of samples into distinct levels of psychological discomfort. 2023 IEEE. -
A Deep Learning Methodology CNN-ADAM for the Prediction of PCOS from Text Report
Text categorization is a popular piece of work in natural language processing (NLP) and machine learning, and Convolutional Neural Networks (CNNs) can be used effectively for this purpose. Although CNNs are traditionally associated with computer vision tasks, they have been adapted and applied successfully to text classification problems. In the proposed study Convolutional Neural Networks (CNNs) with adam optimization algorithm plays a crucial role in detecting PCOS words from sonographic text reports. 2023 IEEE. -
Decoupling Identification Method of Continuous Working Conditions of Diesel Engines Based on a Graph Self-Attention Network
For diesel engine malfunction detection, machine learning-based intelligent detection approaches have made great strides, but some performance deterioration is also observed due to the significant ambient noise and the change in operating circumstances in the actual application situations. Diesel engine fault diagnostic models can be negatively impacted by complex and erratic working circumstances. Identifying the working condition can provide as a baseline for the current unit operating state, which is crucial information when trying to pinpoint the source of an issue. Many existing techniques for identifying operational states use power as an identifier, segmenting it into discrete intervals from which the current state's power may be derived using a classification model. However, the working condition characteristics should be constant, and defining it exclusively in terms of power would lead to the connection of speed and load elements. In this study, we offer a regular working situation model that is independent of speed and load characteristics, and we use a graph self-attention network to construct a model for identifying the working condition. On a diesel engine research bench, a vast amount of experimental data is acquired for training and testing models, including 32 different operating situations under normal and typical fault scenarios. The R2 adj values of 99.70% and 99.27% for normal and typical defect experimental data, correspondingly, demonstrate the efficacy of the suggested technique under the circumstance of uninformed nnerating situations. 2023 IEEE. -
Privacy Optimization in Sensors Based Networks With Industrial Processes Management
The Internet of Things (IoT) also known as IoT has the potential that is required to revolutionize industries, this has been discussed in this research article. Advancements in technology have made devices affordable, efficient and reliable. Different sectors have already started to incorporate these devices into their operations to boost productivity, to minimize failure and downtime. They also use it to optimize resource utilization which is also an important factor. However, the use of these devices also has some security challenges which need to be handled. This research paper proposes a security model specifically designed for process management in the industries. The goal of this model is to find the vulnerabilities, to minimize the risks and threats. Also ensuring integrity, confidentiality and availability of processes is a part of the goal. This paper gives evidence from its implementation and trial apart from its explanation. During the implementation phase, the sensitive data achieved a 100% encryption rate, for protection. Also, integrity checks were conducted on 99.8% of data to guarantee data integrity. 2023 IEEE. -
A Hybrid Edge-Cloud Computing Approach for Energy-Efficient Surveillance Using Deep Reinforcement Learning
This paper explores the novel application of Deep Reinforcement Learning (DRL) in designing a more efficient, scalable, and distributed surveillance architecture, which addresses concerns such as data storage limitations, latency, event detection accuracy, and significant energy consumption in cloud data centers. The proposed architecture employs edge and cloud computing to optimize video data processing and energy usage. The study further investigates the energy consumption patterns of such a system in detail. The implementation leverages machine learning models to identify optimal policies based on system interactions. The proposed solution is tested over an extensive period, resulting in a system capable of reducing latency, enhancing event detection accuracy, and minimizing energy consumption. 2023 IEEE. -
A Comparative Analysis of Autonomous Ledger Systems for Enhanced Blockchain Computing Applications
Due to its potential to completely transform a number of sectors thanks to its irreversible and decentralized ledger system, the use of blockchain technology has recently attracted a lot of interest. Blockchain-based systems still face considerable issues with regard to scalability and effectiveness. This study compares autonomous ledger systems and examines how they are used in blockchain technology computation. With their capacity for self-management and resource allocation optimization, autonomous ledger systems provide intriguing answers to these problems. As examples of autonomous ledger systems, we look at self-care networks, adaptable consensus techniques, and autonomous government systems. We compare them and assess how well they function to improve the speed, security, and scalability of blockchain networks. We also examine the real-world uses of these independent ledger systems in industries including logistics management, banking, and medical services. With the goal of advancing blockchain computing and enabling more reliable and effective decentralized applications, this study intends to shed light on the possibilities of autonomous ledger systems. 2023 IEEE.