Browse Items (2150 total)
Sort by:
-
Recognition of Green Colour Vegetables' Images Using an Artificial Neural Network
Image processing is used in all the domains including agriculture. In this paper, we have introduced a computationally simple and small feature vector, as a tool for the recognition of green colour vegetable images. The RGB colour system is used and the feature set is computationally economic and performs well on locally available vegetable images. For recognition of vegetable images, an ANN-based classifier is deployed. The recognition percentage is in the scale of 74-100 for 15 vegetable types. This work finds application in the packing of vegetables, food processing, automatic vending. 2019 IEEE. -
Non-Contact Vital Prediction Using rPPG Signals
In this paper, we present the clinical significance of various cardiac symptoms with the use of heart rate detection, ongoing monitoring and present emotions. The development of algorithms for remote photoplethysmography has drawn a lot of interest during the past decade (rPPG). As a result, using data gathered from the video feed, we can now precisely follow the heart rate of individuals who are still seated. rPPG algorithms have also been developed, in addition to technique based on hand-crafted characteristics. Deep learning techniques often need a lot of data to train on, but biomedical data frequently lacks real-world examples. The experiment described in this work, we looked at how illumination affected the rPPG signals' SNR. The findings show that the SNR in each RGB channel varies depending on the colour of the light source. Paper describes development in video filtering for recognising the comprehending human face emotions. In our method, emotions are deduced by identifying facial landmarks and analysing their placement. 2023 IEEE. -
Covid-19 Classification and Detection Model using Deep Learning
One of the deadly diseases in recent years is covid19 which is affecting the lives of peoples. Also leading to severe adverse problems and death. Prevention is done using early diagnosis and medication which in turn helps in early detection of the disease. The basic aim of the paper is to identify and further classify the patients using the chest x-rays. From scratch the Convolutional Neural Network is diagnosed producing a very high accurate and optimum results. In recent years, researchers found out that in the radiological images such as like x-rays, the traces of covid-19 can be found. In few areas, a good accuracy of the covid-19 detection cannot be achieved due to lack of the people who test so the artificial intelligence is combined with the radiological image. In machine learning the models used are deep learning by automatizing the actions and making it certain by swift, skillful and proficient outcome produced by the chest images provided by the patients. There are several layers like convolutional layer, max pooling layer etc. which are initiated and are used with aid of ReLU activation function. These images given as inputs are also classified accordingly. There is a sequence of neurons being given as input to the active dense layer and there is a result to the input by a sigmoidal function. There is a rise in efficiency because the models are trained and there is a decline of loss at the same time. If there is a model where fitting is done earlier to the overfitting and is restricted from implementing in the data augmentation. There is a better and efficient involvement of suggestions to models of deep learning. Further there is a classification of chest images for identifying and analyzing covid19. So, to check the Covid detection, the images are used as raw. In this paper a model is proposed to have good accuracy in the classification between Covid and normal and further it can be classified into three categories like Covid, pneumonia, normal. There is a 98.08% for the first one and 87.02% for the second one. By introducing 17 convolutional layers and using the Darknet model used for classifying you only look once (YOLO) for the live identification of the objects and multiple layers of filters are used. In the model there is an initial screening. 2022 IEEE. -
IoT and AI for Real-Time Customer Behavior Analysis in Digital Banking
Digital transformation has revolutionized the banking industry, ushering in an era of enhanced customer experiences and operational efficiency. The convergence of Internet of Things (IoT) and Artificial Intelligence (AI) technologies has further propelled this evolution by providing real-time insights into customer behavior. This research explores the integration of IoT and AI for real-time customer behavior analysis in the context of digital banking. The proliferation of connected devices, ranging from smart phones to wearables, has generated an unprecedented volume of data. IoT facilitates the collection of diverse data points, such as transaction history, location information, and device interactions, creating a comprehensive digital footprint for each customer. Simultaneously, AI algorithms leverage this wealth of data to analyze, predict, and respond to customer behavior dynamically. In the realm of digital banking, understanding and adapting to customer behavior in real-time is crucial for providing personalized services, preventing fraud, and optimizing operational processes. This research delves into the mechanisms by which IoT sensors and devices, coupled with AI algorithms, enable banks to gain deeper insights into customer behavior patterns. Key components of the proposed system include data acquisition through IoT devices, secure data transmission protocols, and AI-driven analytics engines. In conclusion, this research advocates for the symbiotic relationship between IoT and AI in digital banking to enable real-time customer behavior analysis. 2024 IEEE. -
Extractive Text Summarization Using Sentence Ranking
Automatic Text summarization is the technique to identify the most useful and necessary information in a text. It has two approaches 1)Abstractive text summarization and 2)Extractive text summarization. An extractive text summarization means an important information or sentence are extracted from the given text file or original document. In this paper, a novel statistical method to perform an extractive text summarization on single document is demonstrated. The method extraction of sentences, which gives the idea of the input text in a short form, is presented. Sentences are ranked by assigning weights and they are ranked based on their weights. Highly ranked sentences are extracted from the input document so it extracts important sentences which directs to a high-quality summary of the input document and store summary as audio. 2019 IEEE. -
Abusive Words Detection on Reddit Comments Using Machine Learning Algorithms
Utilization of artificial intelligence contributes to the efficient examination of emotions, resulting in valuable insights into the psychological condition of users on a large scale. In this research endeavor, sentiment analysis is conducted on a dataset from Reddit, which was obtained through Kaggle. The feedback in this collection of data was divided into downbeat, neutral, and upbeat sentiments. Various machine learning techniques, like Random Forest, Extreme Gradient Boosting Classifier (XGB), Gradient Boosting Machine (GBM), Support Vector Machine (SVM), and Convolutional Neural Network (CNN), were detected and examined to assess their effectiveness in sentiment classification. The review of these techniques comprised performance criteria such as F1 Score, accuracy, precision, and recall. Additionally, confusion matrices were utilized to assess the algorithms' proficiency in identifying abusive language. The investigation's conclusions indicate that, when it comes to sentiment analysis, the random forest method performs better than any other strategy, with a maximum accuracy of 0.99 that is on par with the CNN model's accuracy of 0.98. Moreover, random forest proves to be the most effective algorithm in recognizing negative comments and abusive language. This study underscores the significance of employing machine learning algorithms in sentiment analysis, content moderation, social media monitoring, and customer feedback analysis, emphasizing their role in enhancing automated systems that aim to comprehend user sentiments in online discussions. 2024 IEEE. -
The Road to Reducing Vehicle CO2 Emissions: A Comprehensive Data Analysis
In recent years, the influence of carbon dioxide (CO2) releases on the environment have become a major concern. Vehicles are one of the major sources of CO2 emissions, and their contribution to climate change cannot be ignored. This research paper aims to investigate the CO2 emissions of vehicles and compare them with different types of engines, fuel types, and vehicle models. The study was carried out by gathering information about the CO2 emissions of vehicles from the official open data website of the Canadian government. Data from a 7-year period are included in the dataset, which is a compiled version. There is a total of 220 cases and 9 variables. The data is analyzed using statistical methods and tests to identify the significant differences in CO2 emissions among different Car Models. The results indicate that vehicles with diesel engines emit higher levels of CO2 compared to those with gasoline engines. Electric vehicles, on the other hand, have zero CO2 emissions, making them the most environmentally friendly option. Furthermore, the study found that the CO2 emissions of vehicles vary depending on the type of fuel used. The study also reveals that the CO2 emissions of vehicles depend on the model and age of the vehicle. Newer models tend to emit lower levels of CO2 compared to older models. In conclusion, this study provides valuable insights into the CO2 emissions of Cars and highlights the need to adopt cleaner and more sustainable transportation options. The Author(s), under exclusive license to Springer Nature Switzerland AG 2024. -
CNN-Bidirectional LSTM based Approach for Financial Fraud Detection and Prevention System
Detecting fraudulent activity has become a pressing issue in the ever-expanding realm of financial services, which is vital to ensuring a positive ecosystem for everyone involved. Traditional approaches to fraud detection typically rely on rule-based algorithms or manually pick a subset of attributes to perform prediction. Yet, users have complex interactions and always display a wealth of information when using financial services. These data provide a sizable Multiview network that is underutilized by standard approaches. The proposed method solves this problem by first cleaning and normalizing the data, then using Kernel principal component analysis to extract features, and finally using these features to train a model with CNN-BiLS TM, a neural network architecture that combines the best parts of the Bidirectional Long Short-Term Memory (BiLS TM) network and the Convolution Neural Network (CNN). BiLSTM makes better use of how text fits into time by looking at both the historical context and the context of what came after. 2023 IEEE. -
Interpreting the Evidence on Life Cycle to Improve Educational Outcomes of Students Based on Generalized ARC-GRU Approach
Research on the effects of teachers' fatigue on students' learning has been significantly less common than research on the effects of teachers' fatigue on teachers' own performance. Therefore, the purpose of this research is to see if teachers' emotional weariness has any bearing on their students' performance in the classroom. Consideration is given to a student's grades and their impressions of whether or not the system receive assistance from teachers, as well as to the student's general outlook on school, confidence in their own abilities, and faith in the availability of faculty support. Data preparation, feature extraction, and model training are the first steps in the proposed approach. Indicators of the quality of the education being provided are eliminated (by outlier removal and feature scaling). k-mean clustering approach is a technique of clustering which is commonly used in feature extraction. Following feature extraction, GARCH-GRU models are trained. The proposed approach is superior to two popular alternatives, ARCH and GRU. Using the provided method, the system were able to achieve a maximum accuracy of 97.07%. 2024 IEEE. -
Enhanced Social Media Profile Authenticity Detection Using Machine Learning Models and Artificial Neural Networks
Fake engagement is one of the main issues with online networks or ONSs, which are used to artificially boost an account's popularity, this study examines the effectiveness of seven sophisticated Machine Learning Algorithms, Random Forest Classifier, Decision Tree Classifier, XGBoost, LightGBM, Extra Trees Classifier, and SVM, and got 93% accuracy in Decision Tree Classifier. In order to solve overfitting issues and improve model resilience, the paper proposes Generative Adversarial Networks (GANs) and uses K-Fold Cross-Validation. Furthermore, design a Gan-ANN model that combines Batch Normalization and Artificial Neural Networks (ANN) with GAN-generated synthetic data is investigated. The enhanced dataset seeks to strengthen model performance and generalization when combined with cutting-edge modeling methods. This study aims to improve model scalability, predictive accuracy, and dependability across different machine learning paradigms. 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. -
LBP-GLZM Based Hybrid Model for Classification of Breast Cancer
Classifying mammogram images is difficult because of their complex backgrounds and the differences in resolutions across the images. One of the toughest parts is telling the difference between harmless (benign) and harmful (malignant) tissue. This is hard because the differences between them are incredibly subtle. As a consequence, the distinctive features embedded within tissue patches become not just relevant but critical for the accurate and automatic classification of these images. Traditionally, efforts to automate this classification process have encountered limitations when relying on a singular feature or a restricted set of characteristics. The subtle variations in texture within these images often render such approaches insufficient in achieving high-quality categorization results. Recognizing this, the present investigation undertakes a more comprehensive approach by incorporating distinct feature extraction techniques - specifically, the utilization of Local Binary Pattern (LBP) and Gray Level Zone Matrix (GLZM). These techniques are adept at capturing and delineating the nuanced texture features inherent in mammogram images. By extracting and analyzing these textural nuances, the aim is to construct a hybrid model capable of classifying mammograms into three distinct categories: malignant, benign, and without the necessity for further examination or follow-up. This proposed hybrid model holds significant promise in the field of mammography classification by leveraging the strengths and complementary attributes of multiple feature extraction methods. The integration of LBP and GLZM aims not only to enhance the accuracy of classification but also to improve the robustness of the system in identifying subtle yet crucial differences in tissue textures. Ultimately, the goal is to create a hybrid feature extraction framework that augments the diagnostic capabilities of mammography, providing more precise and reliable categorization of breast tissue for effective medical decision-making and patient care. 2024 IEEE. -
Analysis of error rate for various attributes to obtain the optimal decision tree
The competitiveness and computational intelligence are required to increase the gross profit of the product in a market. The classification algorithm rpart is applied on retail market dataset. The regression rpart decision tree algorithm is implemented with principal component analysis to impute data in the missing part of the dataset. The objective is to obtain an optimal tree by analysing cross validation error, standard deviation error, and number of splits and relative error of various attributes. The results of various attributes by ANOVA method are compared to choose the best optimal tree. The tree with minimum error rate is considered for the optimal tree. Copyright 2022 Inderscience Enterprises Ltd. -
Evalutionary compting model for QoS provisioning in WiMAX
In recent time wireless technology is adopted widely for connecting remote user over network. WiMAX is an attractive technology for provisioning high data rate connectivity and coverage. QoS is a required parameter for analyzing the system performance. Before allocation of the bandwidth in the network, physical layer information is required for improving QoS performance. Many modulation techniques are used in WiMAX network. An adaptive approach is required for selection of modulation scheme to maximize network performance. Physical layer information is used for selection of modulation scheme. An adaptive genetic based scheduling is proposed in this paper for improved QoS. Experiments are conducted to evaluate performance of proposed approach in term of throughput, successful transmission and packet collision over existing approach. The outcome shows significant performance over existing approach. 2017 IEEE. -
Convolutional Neural Networks for Automated Detection and Classification of Plasmodium Species in Thin Blood Smear Images
There has been a continued transmission of malaria throughout the world due to protozoan parasites from the Plasmodium species. As for treatment and control, it is very important to make correct and more efficient diagnostic. In order to observe the efficiency of the proposed approach, This Research built a Convolutional Neural Network (CNN) model for Automated detection and classification on thin blood smear images of Plasmodium species. This model was built on a corpus of 27558 images, included five Plasmodium species. Our CNN model got an overall accuracy of 96% for the cheating detection with an F 1score of 0.94. In the detection of the presence of malaria parasites the test accuracy conducted was as follows: 8%. Species-specific classification accuracies were: P. falciparum (95.7%), P. vivax (94.9%), P. ovale (93.2%), P. malaria (92.8%) and P. Knowles (91, 5%). As for the model SL was found to have sensitivity of 97.3% And the specificity in this case is 9 6. 1 %. The proposed CNN-based approach provides a sound and fully automated solution for malarial parasite detection and species determination, which could lead to better diagnostic performances in day-to-day practices. 2024 IEEE. -
Comparative Analysis of Machine Learning Models in Predicting Academic Outcomes: Insights and Implications for Educational Data Analytics
In the evolving landscape of educational research, the predictive analysis of student performance using data science has garnered significant interest. This study investigates the influence of diverse factors on academic outcomes, ranging from personal demographics to socioeconomic conditions, to enhance educational strategies and support mechanisms. We employed a diverse ml models to analyze a information containing academic records and socioeconomic information. The models tested include Logistic Regression, Random Forest (RF), Gradient Boosting (GB), Support Vector Machines (SVC), K-Nearest Neighbors (KNN), Gaussian Naive Bayes, and Decision Trees. The process involved comprehensive data preprocessing, exploratory analysis, model training, and evaluation based on metrics such as precision, recall, accuracy, and F1 score. The results indicate that ensemble methods, specifically RF and GB, demonstrate superior efficacy in accurately predicting categories of student performance such as 'Enrolled,' 'Graduated,' and 'Dropped Out.' These models excelled in handling the complex interplay of varied predictors affecting student success. The results further underline the potential of advanced ensemble ML techniques in significantly outperforming the prediction accuracy in the academic domain, hence facilitating the tailoring of educational interventions to foster improved engagement and better outcomes for students. This has provided a comparative analysis of the methods that guide the future application of predictive analytics in education. 2024 IEEE. -
Towards an Improved Model for Stability Score Prediction: Harnessing Machine Learning in National Stability Forecasting
In our increasingly interconnected world, national stability holds immense significance, impacting global economics, politics, and security. This study leverages machine learning to forecast stability scores, essential for understanding the intricate dynamics of country stability. By evaluating various regression models, our research aims to identify the most effective methods for predicting these scores, thus deepening our insight into the determinants of national stability. The field of machine learning has seen remarkable progress, with regression models ranging from conventional Linear Regression (LR) to more complex algorithms like Support Vector Regression (SVR), Random Forest (RF), and Gradient Boosting (GB). Each model has distinct strengths and weaknesses, necessitating a comparative analysis to determine the most suitable model for specific predictive tasks. Our methodology involves a comparative examination of models such as LR, Polynomial Regression (PR), Lasso, Ridge, Elastic Net (ENR), Decision Tree (DT), RF, GB, K-Nearest Neighbors (KNN), and SVR. Performance metrics like Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), Mean Squared Error (MSE), and R-squared (R2) assess each model's predictive accuracy using a diverse dataset of country stability indicators. This study's comprehensive model comparison adds novelty to predictive analytics literature. Our findings reveal significant variations in the performance of different regression models, with certain models exhibiting exceptional predictive accuracy, as indicated by high R2 values and low error metrics. Notably, models such as LR, SVR, and Elastic Net demonstrate outstanding performance, suggesting their suitability for stability score prediction. 2024 IEEE. -
Prediction of Stock Prices using Prophet Model with Hyperparameters tuning
As part of the data analytical process, predicting and time - series are crucial. In academics and financial research, anticipating share prices is a prominent and significant subject. A share market would be an uncontrolled environment for anticipating shares since there are no fundamental guidelines for evaluating or anticipating share prices there. As a result, forecasting share prices is a difficult time-series issue. fundamental, technical, time series predictions and analytical strategies are just a few of the various techniques and approaches that machine learning uses to execute stock value predictions. This article implements the stock price prediction, Researchers compared the model of the prophet with the tuned model of the prophet. By utilizing the tuning of hyperparameters using parameter grid search to improve the performance of the model accuracy for the best prediction. The findings of the study demonstrated that tuned model of the prophet with hyperparameters tuning which results in model accuracy and based on the experimental findings mean squared error (MSE) and mean absolute percentage error (MAPE) has significant improvement. 2022 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. -
A Hybrid Approach for Predictive Maintenance Monitoring of Aircraft Engines
The realm of aircraft maintenance involves predictive maintenance, which utilizes historical data and machine parts' performance to anticipate the need for maintenance activities. The primary focus of this paper is to delve into the application of predictive maintenance of aircraft gas turbine engines. Our methodology involves assigning a randomly chosen deterioration value and monitoring the change in flow and efficiency over time. By carefully analyzing these factors, we can deduce whether the engines are at fault and whether their condition will deteriorate further. The ultimate objective is to identify potential engine malfunctions early to prevent future accidents. Recent years have witnessed the emergence of multiple machine learning and deep learning algorithms to predict the Remaining Useful Life (RUL) of engines. The precision and accuracy of these algorithms in assessing the performance of aircraft engines are pretty promising. We have incorporated a hybrid model on various time series cycles to enhance their efficacy further. Employing data collected from 21 sensors, we can predict the remaining useful life of the turbine engines with greater precision and accuracy. 2024 IEEE.