Browse Items (11809 total)
Sort by:
-
Comparative Performance Analysis of Machine Learning and Deep Learning Techniques in Pneumonia Detection: A Study
Pneumonia is a bacterial or viral infection that inflames the air sacs in one or both lungs. It is a severe life-threatening disease, making it increasingly necessary to develop accurate and reliable artificial intelligence diagnosis models and take early action. This paper evaluates and compares various Machine Learning and Deep Learning models for pneumonia detection using chest X-rays. Six machine learning models -Logistic Regression, KNN, Decision Tree, Random Forest, Naive Bayes, and Support Vector Machines - and three deep learning models - CNN, VGG16, and ResNet - were created and compared with each other. The results exhibit how just the model choice can significantly affect the quality and inerrancy of the final diagnostic tool. 2023 IEEE. -
Weather Forecasting Accuracy Enhancement Using Random Forests Algorithm
In today's world, weather forecasting is essential for decision-making in a variety of fields, including agriculture, transportation, and disaster preparedness. It's not simple to make weather predictions. Today, both in business and academia, data analytics is growing in importance as a tool for decision-making. The adoption of data-driven concepts is for our graduates, enhancing their marketability. Data Analytics us a study belonging to science that analyses gathered raw data, which makes conclusions about the particular information. Data analytics has been used by many sectors recently, such as hospitality, where this industry can collect data, find out where the problem is, and manage to fix the problem. Nominal, ordinal, interval, and ratio data levels are the four types of data measurement. Applications of data analytics can be found in many industries, including shipping and logistics, manufacturing, security, education, healthcare, and web development. Any business that wants to succeed in the modern digital economy should make analytics a core focus. To make such data meaningful, a transformation engine was used with types from several sources. Ironically, this has made analytics harder for businesses. As businesses employ more platforms and applications, the amount of data available has grown tremendously. This article focuses on different applications of data analytics in the modern world. Weather forecasting is a highly intricate and multifaceted process that draws upon data from various sources. It relies on a combination of scientific studies and sophisticated weather models to decipher the vast amount of information available. 2023 IEEE. -
Machine Learning Based Crime Identification System using Data Analytics
Poverty is known to be the mother of all crimes, and a vast percentage of people in India live below the poverty line. In India, the crime rate is rapidly rising. The police officers must spend a significant amount of time and personnel to identify suspects and criminals using current crime investigation. In this research, the method presented for designing and implementing crime identification and criminal recognition systems for Indian metropolitans is utilizing techniques of data mining. These occurrences are represented by 35 predefined crime attributes. Access to the crime database is protected by safeguards. The pending four subjects are important for crime unmasking, identification and estimation of criminals, and crime authentication, in that order. The detection of crime is investigated with the help of K-Means clustering, which iteratively builds two crime batches based on congruent criminal features. Google Maps is to enhance the k-means visualization. K-Nearest Neighbor classification is used to examine criminal identification and forecasting. This is used for the authentication of the results. The technique benefits society by helping investigative authorities in crime solving and criminal recognition, resulting in lower crime rates. This research study describes a way for creating and deploying crime solving and criminal recognition systems for Indian metro's using data mining tools in this study. The method consists of data evulsion, data pre- processing, clustering, Google map delegation and classification. The first module, data evulsion, retrieves unformed or unrecorded crime datasets from several criminal sources online from 2000 to 2012. In the second module, Data pre-processing cleans, assimilates, and reduces the obtained criminal data into organized 5,038 crime occurrences. Several predefined criminal traits represent these instances. Safeguards are in place to prevent unauthorized access to the crime index. The remaining components are critical for detecting crimes, criminal identity and prediction, and crime verification, in that sequence. The investigation of crimes is investigated using k-means clustering, which gives results repeatedly. 2023 IEEE. -
Resource Aware Weighted Least Connection Load Balancing Technique in Cloud Computing
Cloud computing became a pivotal for the most of the real time applications. In cloud computing, the customer demands the services with the best performance even when the application is expanding rapidly. Therefore, it is essential to manage the resources effectively because the number of users and services growing proportionately. The main aim of the load balancing technique is to allocate the customers' requests with the large pool of resources efficiently. The problem is how to evenly distribute the load of requests among the compute nodes according to their capacity. Therefore, there is a need for an effective load balancing technique for smooth continuity of operations in a distributed environment with a heterogeneous server configuration. This paper presents a novel load balancing technique, namely, Resource aware weighted least connection load balancing which addresses the above said problem efficiently. The essence of this work is to assign the requests across multiple servers based on the requested resource and the status of the number of connections presently served by each server. This work used standard score technique to enumerate the weight of each node. Experiments were conducted using Cloud Analyst, a famous cloud simulator breed from CloudSim. Appropriate performance parameters were analysed to measure the effectiveness of the proposed technique. Future directions for the extension of the implemented technique also identified. 2023 IEEE. -
Martian Habitats: A Review
Establishing colonies in Lunar and Martian environments is the major task of our primary means to become a multi-planetary civilization. The Space Exploration Initiative (SEI), administered by President George H.W. Bush in 1989, was the first spark that ignited humanity's vision to establish space settlements beyond low Earth orbit (LEO) (Marc M. Cohen, 2015). At present, private space companies (like SpaceX and Blue Origin) are competing to be the first ones to colonise space. From the late 1980s to the present space race, many space habitat designs to suit human factors, ensure protection from space radiation, and be capable of regulating our day-to-day activities have been proposed for both lunar and martian settlements, respectively. In this paper, only Martian settlements are focused, and the reason for that follows next. While the moon is closer to Earth than Mars, Mars has several other advantages that make it an equal, if not a better candidate for colonisation. Some of the reasons why martian colonisation is preferred over lunar colonization include the presence of an atmosphere on Mars, its resource-rich nature, and its rotation period being closer to Earth's rotation period (Mars has 24.5 hours per day, while the moon has 28-day days) (Kamrun Narher Tithi, 2017). Another added advantage is its proximity to the main belt asteroids, which will further increase the potential for space mining in the future. So this paper will be a review of the various Martian habitat designs proposed over the last one and a half decades in terms of their designs, construction and challenges. To do so, it is assumed that every step associated with delivering the habitats to the Martian environment is achievable. These steps include the following: propulsion systems for long-term spaceflights; launch vehicles capable of lifting the habitats and fitting the habitat modules within them (Marc M. Cohen, 2015). Copyright 2023 by the International Astronautical Federation (IAF). All rights reserved. -
Streamlined Deployment and Monitoring of Cloud-Native Applications on AWS with Kubernetes Prometheus Grafana
As organizations increasingly move their applications to the cloud, it becomes essential to have an efficient and cost-effective method for deploying and managing those applications. Manual deployment can be time-consuming, error-prone, and expensive. Additionally, managing logs and monitoring resources for each deployment can lead to even greater costs. To address these challenges, we propose implementing an automation strategy for deployment in the cloud. With automation, the deployment process can be streamlined and standardized across different cloud providers, reducing the potential for errors and saving time and resources. Furthermore, a central log system can be implemented to manage logs from different deployments in one location. This provides a unified view of all logs and allows for easier troubleshooting and analysis. Automation can also be used to set up monitoring resources, such as alerts and dashboards, across different deployments. Overall, implementing an automation strategy for deployment in the cloud can help organizations save time and resources while improving their ability to manage and monitor their applications. A centralized log management system can further enhance these benefits by providing a unified view of logs from all deployments 2023 IEEE. -
An Empirical and Statistical Analysis of Regression Algorithms Used for Mental Fitness Prediction
In today's focus on mental well-being, technology's capability to predict and comprehend mental fitness holds substantial significance. This study delves into the relationship between mental health indicators and mental fitness levels through diverse machine learning algorithms. Drawing from a vast dataset spanning countries and years, the research unveils concealed patterns shaping mental well-being. Precise analysis of key mental health conditions reveals their prevalence and interactions across demographics. Enriched by insights into Disability-Adjusted Life Years (DALYs), the dataset offers a comprehensive view of mental health's broader impact. Through rigorous comparative analysis, algorithms like Linear Regression, Random Forest, Support Vector Regression, Gradient Boosting, K-nearest neighbors and Theil Sen Regression are assessed for predictive accuracy. Mean squared error (MSE), root mean squared error (RMSE), and Rsquared (R2) scores are used to assess the predictive accuracy of each algorithm. Results show that Mean Squared Error (MSE) ranged from 0.030 to 1.277, Root Mean Squared Error (RMSE) from 0.236 to 1.130, and R-squared (R2) scores ranged between 0.734 and 0.993, with Random Forest Regressor achieving the highest accuracy. This study offers precise prognostications regarding mental fitness and establishes the underpinnings for the creation of effective tracking tools. Amidst society's endeavor to tackle intricate issues surrounding mental health, our research facilitates well-informed interventions and individualized strategies. This underscores the noteworthy contribution of technology in shaping a more Invigorating trajectory for the future. 2023 IEEE. -
Object Detection with Augmented Reality
This study describes an artificial intelligence (AI)-based object identification system for detecting real-world items and superimposing digital information in Augmented Reality (AR) settings. The system evaluates the camera stream from an AR device for real-Time recognition using deep learning algorithms trained on a collection of real-world items and their related digital information. Object recognition applications in AR include gaming, education, and marketing, which provide immersive experiences, interactive learning, and better product presentations, respectively. However, challenges such as acquiring larger and more diverse datasets, developing robust deep learning algorithms for varying conditions, and optimizing performance on resource-constrained devices remain. The AI-based object recognition system demonstrates the potential to transform AR experiences across domains, while emphasizing the need for ongoing research and development to fully realize its capabilities. 2023 IEEE. -
Deep Learning for Arrhythmia Classification: A Comparative Study on Different Deep Learning Models
Arrhythmias, or irregular heart rhythms, are a major global health concern. Since arrhythmias can cause fatal conditions like cardiac failure and strokes, they must be rapidly identified and treated. Traditional arrhythmia diagnostic techniques include manual electrocardiogram (ECG) image interpretation, which is time consuming and frequently required for expertise. This research automates and improves the identification of heart problems, with a focus on arrhythmias, by utilizing the capabilities of deep learning, an advanced machine learning technique that performs well at recognizing patterns in data. Specifically, we implement and compare Custom CNN, VGG19, and Inception V3 deep learning models, which classify ECG images into six categories, including normal heart rhythms and various types of arrhythmias. The VGG19 model excelled, achieving a training accuracy of 95.7% and a testing accuracy of 93.8%, showing the effectiveness of deep learning in the comprehensive diagnosis of heart diseases. 2023 IEEE. -
An Early-Stage Diabetes Symptoms Detection Prototype using Ensemble Learning
Diabetes is one of the most increasing health issues that the whole world is facing. Recent research has shown that diabetes is spreading quickly in India. Having more than 77 million sufferers, India is actually regarded as the diabetes capital of the world. The lifestyle and eating patterns of people who move from rural to urban settings alter, which raises the prevalence of diabetes. Diabetes has been linked to consequences like vision loss, renal failure, nerve damage, cardiovascular disease, foot ulcers, and digestive issues. Diabetes can harm the blood arteries and neurons in a variety of organs. FPG (Flaccid Plasma Glucose) is a popular test that is done to find out whether a person is a diabetic patient or not. However, not all people consistently take medication and neither monitor their blood sugar levels on a regular basis. Early detection of this disease is also an important thing that people usually don't do. Technology these days has emerged a lot in the healthcare zone. Many prototypes have already been made for the detection of diabetes. The prototype discussed in this paper is an ensemble learning approach for the detection of diabetes in a very early stage. Ensemble learning which includes the use of multiple model prediction has been used to make the outcome stronger and more trustworthy. The overall accuracy achieved by the model is 96.54%. XGBoost also records the minimal execution time of 2.77 seconds only. 2023 IEEE. -
Sustainable Technologies for Recycling Process of Batteries in Electric Vehicles
The effective management of batteries has always been a key concern for people because of the imposing challenges posed by battery waste on the environment. This paper explores strategic perspectives on the sustainable management of batteries incorporating modern techniques and scientific methodologies giving batteries a second-life application. A paradigm shift towards the legitimate use of the batteries by the introduction of round economy for battery materials and simultaneously checking the biological impression of this fundamental innovation area. 2023 IEEE. -
An Improved Image Up-Scaling Technique using Optimize Filter and Iterative Gradient Method
In numerous realtime applications, image upscaling often relies on several polynomial techniques to reduce computational complexity. However, in high-resolution (HR) images, such polynomial interpolation can lead to blurring artifacts due to edge degradation. Similarly, various edge-directed and learning-based systems can cause similar blurring effects in high-frequency images. To mitigate these issues, directional filtering is employed post corner averaging interpolation, involving two passes to complete the corner average process. The initial step in low-resolution (LR) picture interpolation involves corner pixel refinement after averaging interpolation. A directional filter is then applied to preserve the edges of the interpolated image. This process yields two distinct outputs: the base image and the detail image. Furthermore, an additional cuckoo-optimized filter is implemented on the base image, focusing on texture features and boundary edges to recover neighboring boundary edges. Additionally, a Laplacian filter is utilized to enhance intra-region information within the detailed image. To minimize reconstruction errors, an iterative gradient approach combines the optimally filtered image with the sharpened detail image, generating an enhanced HR image. Empirical data supports the effectiveness of the proposed algorithm, indicating superior performance compared to state-of-the-art methods in terms of both visual appeal and measured parameters. The proposed method's superiority is demonstrated experimentally across multiple image datasets, with higher PSNR, SSIM, and FSIM values indicating better image degradation reduction, improved edge preservation, and superior restoration capabilities, particularly when upscaling High-Frequency regions of images. 2023 IEEE. -
Students Perception of Chat GPT
An artificial intelligence based Chatbot, ChatGPT was launched by Open AI in November 2022. In the field of education, ChatGPT has several benefits as well as challenges. Chat GPT can be considered as an advanced and a powerful tool to enhance the learning experience. It adds value to the education system only when it is used wisely. However, it is important to understand that the challenges must be addressed. It may act as a good source for collating the information, but it is always advised by the researchers that ones own perspective must be added to draw inferences from the output generated by ChatGPT. Our study supports the finding that ChatGPT can be used for the generation of ideas or to learn a new language. It also becomes imperative for the faculties to motivate students to use ChatGPT and add their inferences as well. AI models like ChatGPT can provide assistance, answer questions and provide explanations on various topics, making learning more accessible and tailored to individual needs. With this paper, we aim to provide a more informed discussion around the usage of ChatGPT in education. 2023 IEEE. -
Machine Learning Based Spam E-Mail Detection Using Logistic Regression Algorithm
The rise of spam mail, or junk mail, has emerged as a significant nuisance in the modern digital landscape. This surge not only inundates user's email inboxes but also exposes them to security threats, including malicious content and phishing attempts. To tackle this escalating problem, the proposed machine learning-based strategy that employs Logistic Regression for accurate spam mail prediction. This research is creating an effective and precise spam classification model that effectively discerns between legitimate and spam emails. To achieve this, we harness a meticulously labeled dataset of emails, each classified as either spam or non-spam. This model is to apply preprocessing techniques to extract pertinent features from the email content, encompassing word frequencies, email header data, and other pertinent textual attributes. The choice of Logistic Regression as the foundational classification algorithm is rooted in its simplicity, ease of interpretation, and appropriateness for binary classification tasks. To process train the model using the annotated dataset, refining its hyper parameters to optimize its performance. By incorporating feature engineering and dimensionality reduction methodologies, bolster the model's capacity to generalize effectively to unseen data. Our evaluation methodology encompasses rigorous experiments and comprehensive performance contrasts with other well-regarded machine learning algorithms tailored for spam classification. The assessment criteria encompass accuracy, precision, recall, and the F1 score, offering a holistic appraisal of the model's efficacy. Furthermore, we scrutinize the model's resilience against diverse forms of spam emails, in addition to its capacity to generalize to new data instances. This model is to findings conclusively demonstrated that our Logistic Regression-driven spam mail prediction model achieves a competitive performance standing when juxtaposed with cutting-edge methodologies. The model adeptly identifies and sieves out spam emails, thereby cultivating a more trustworthy and secure email environment for users. The interpretability of the model lends valuable insights into the pivotal features contributing to spam detection, thereby aiding in the identification of emerging spam patterns. 2023 IEEE. -
A Comprehensive Research on Deep Learning Based Routing Optimization Algorithms in Software Defined Networks
Discovering an optimal routing in Software Defined Networks (SDNs) is challenging due to several factors like scalability issues, interoperability, reliability, poor configuration of controllers and security measures. The compromised SDN controller attacks at the control plane layer, packet losses in the topology and end-to-end delay are the most security risk factors in SDNs. To overcome this, in most of the existing researches, Deep Reinforcement Learning (DRL) algorithm with various optimization techniques was implemented for optimal routing in SDN by providing link weights to balance the end-to-end delay and packet losses. DRL used Deterministic Policy Gradient (DPG) method which acts as an actor-critic reinforcement learning agent that searches for an optimal policy to minimize the expected cumulative long-term reward. However, discovering an optimal routing with efficient security measures is still a major challenge in SDNs. This research proposes a detailed review of routing optimization algorithms in SDN using Deep Learning (DL) methods which supports the researchers in accomplishing a better solution for future research. 2023 IEEE. -
Performance Analysis of Deep Learning Pretrained Image Classifiation Models
Convolutional Neural Networks (CNNs) is revolutionized in the field of computer vision, with the high accuracy and capability to learn features from raw data. In this research work focused on a comparative analysis of two popular CNN architectures, VGG16 and VGG19. The CIFAR dataset consists of 60,000 images, each with a resolution of 32x32 and it's belong to one of the 10 classes. Experimental results are compared with VGG16 and VGG19 in terms of their accuracy and training time, and to identify any differences in their ability to learn features from the CIFAR-10 dataset. The results of this research can aid in directing the choice of appropriate architectures for image classification tasks as well as the advantages of optimisation strategies for enhancing the efficiency of deep learning models. In order to enhance the performance of these structures, more optimisation methods and datasets may be investigated in subsequent research. 2023 IEEE. -
Artificial Intelligence Based Enhanced Virtual Mouse Hand Gesture Tracking Using Yolo Algorithm
Virtual mouse technology has revolutionized human computer interaction, allowing users to interact with digital environments without physical peripherals. The concept traces back to the late 1970s, and over the years, it has evolved with significant advancements in computer vision, motion tracking, and gesture recognition technologies. In recent times, machine learning techniques, particularly YOLOv8, have been integrated into virtual mouse technology, enabling accurate and swift detection of virtual objects and surfaces. This advancement enhances seamless interaction, intuitive hand gestures, and personalized virtual reality experiences tailored to individual user preferences. The proposed model, EHT (Enhanced Hand Tracking), leverages the power of YOLOv8 to address the limitations of existing models, such as Mediapipe. EHT achieves higher accuracy in hand tracking, real-Time hand gesture recognition, and improved multi-user interactions. It adapts to users' unique gestures over time, delivering a more natural and immersive computing experience with accuracy rates exceeding those of Mediapipe. For instance, across multiple sample datasets, EHT consistently outperformed Mediapipe in hand tracking accuracy. In Sample Dataset 1, EHT demonstrated an accuracy of 98.3% compared to Mediapipe's 95.65%. Similarly, in Sample Dataset 2, EHT achieved an accuracy of 99.35%, surpassing Mediapipe's 94.63%. Even in Sample Dataset 3, EHT maintained its superiority with an accuracy of 98.54 %, whereas Mediapipe achieved 98.26%. The successful implementation of EHT requires a custom dataset and optimization techniques to ensure efficiency on virtual reality hardware. EHT model is anticipated redefining how users interact with digital environments, unlocking new possibilities for intuitive and immersive computing experiences. 2023 IEEE. -
Malicious Traffic Classification in WSN using Deep Learning Approaches
Classifying malicious traffic in Wireless Sensor Networks (WSNs) is crucial for maintaining the network's security and dependability. Traditional security techniques are challenging to deploy in WSNs because they comprise tiny, resourceconstrained components with limited processing and energy capabilities. On the other hand, machine learning-based techniques, such as Deep Learning (DL) models like LSTMs, may be used to detect and categorize fraudulent traffic accurately. The classification of malicious traffic in WSNs is crucial because of security. To protect the network's integrity, data, and performance and ensure the system functions properly and securely for its intended use, hostile traffic categorization in WSNs is essential. Classifying malicious communication in a WSN using a Long Short-Term Memory (LSTM) is efficient. WSNs are susceptible to several security risks, such as malicious nodes or traffic that can impair network performance or endanger data integrity. In sequential data processing, LSTM is a Recurrent Neural Network (RNN) appropriate for identifying patterns in network traffic data. 2023 IEEE. -
Deep Learning Based Face Recognized Attendance Management System using Convolutional Neural Network
In today's digital age, manual attendance tracking is plagued by inefficiency and the potential for inaccuracies, often leading to proxy attendance. The main aim of this research work is to manage and monitor the student's attendance by using face recognition technology. This proposed model is mainly categorized four major modules. First module is database creation. Second module is face detection. Then third module is face recognition and final module is automatic attendance updating process. Student images are compiled to create a comprehensive database, ensuring inclusivity across the class roster. The system utilizes the face recognition library, which relies on deep learning based algorithms for face detection and recognition during testing. This face recognition part Convolutional Neural Network algorithm is used. The system matches detected faces with the known database and marks attendance, ensuring a streamlined and accurate attendance tracking process. This innovative approach has the potential to revolutionize attendance management in educational settings, offering a contactless and efficient solution while mitigating proxy attendance concerns. The proposed model is to compare the accuracy level of face recognition. 2023 IEEE. -
Vehicular Propagation Velocity Forecasting Using Open CV
This work presents a predictive learning driven methodology for recognizing the vehicular velocity. The developed model uses machine vision models to trace and detect vehicular movement in timely manner. It further deploys a machine tested framework for estimation of its velocity on basis of the accumulated information. The technique depends upon a CNN model that is validated with a standardized instances of vehicular scans and corresponding velocity parameters. The proposed model generates good efficiency and robustness in determining velocities across test conditions which encompass various kinds of vehicles and lighting scenarios. An optimal vehicular frequency is noted with heavy-weight vehicles in place in comparison to other vehicles. A mean latency period of 1.25 seconds and an error rate of 0.05 is observed with less road traffic in place. The suggested approach can be of great help in transportation systems, traffic monitoring and enhancing road safety. 2023 IEEE.