13264 research outputs found
Sort by
Ensemble-SMOTE: Mitigating Class Imbalance in Graduate on Time Detection
No full textIn education, detecting students graduating on time is difficult due to high data complexity. Researchers have employed various approaches in identifying on-time graduation with Machine Learning, but it remains a challenging task due to the class imbalance in the dataset. This study has aimed to (i) compare various class imbalance treatment methods with different sampling ratios, (ii) propose an ensemble class imbalance treatment method in mitigating the problem of class imbalance, and (iii) develop and evaluate predictive models in identifying the likelihood of students graduating on time during their studies in university. The dataset is collected from 4007 graduates of a university from year 2021 and 2022 with 41 variables. After feature selection, various class imbalance treatment methods were compared with different sampling ratios ranging from 50% to 90%. Moreover, Ensemble-SMOTE is proposed to aggregate the dataset generated by Synthetic Minority Oversampling Technique variants in mitigating the problem of class imbalance effectively. The dataset generated by class imbalance treatment methods were used as the input of the predictive models in detecting on-time graduation. The predictive models were evaluated based on accuracy, precision, recall, F0.5-score, F1-score, F2-score, Area under the Curve, and Area Under the Precision-Recall Curve. Based on the findings, Logistic Regression with Ensemble-SMOTE outperformed other predictive models, and class imbalance treatment methods by achieving the highest average accuracy (87.24), recall (92.50%), F1-score (91.30%), and F2-score (92.02%) from 6th until 10th trimester. To assess the effectiveness of class imbalance treatment methods, Friedman test is performed to determine on significant difference between the models after applying Shapiro-Wilk test in normality test. Consequently, Ensemble-SMOTE is ranked as the top-performers by achieving the lowest value in the average rank based on the performance metrics. Additional research could incorporate and examine more complicated approaches in mitigating class imbalance when the dataset is highly imbalanced
Sine Cosine Algorithm for Enhancing Convergence Rates of Artificial Neural Network: A Comparative Study
No full textArtificial neural networks (ANNs) is
widely adopted by researchers for classification tasks
due to their simplicity and superior performance. This
study offerings the ANN and it variant such as Elman
Neural Network (NN) model to address its strengths,
although it faces with issues like local minima and slow
convergence. This study presents a comprehensive
evaluation of four distinct algorithms for classification
tasks, focusing on their performance on both training
and testing datasets. These algorithms such as Sine
Cosine Algorithm is integrated with Artificial Neural
Networks (SCA_ANN), Back Propagation Neural
Networks (SCA_BP), Elman Neural Networks
(SCA_ElmanNN), and Elman Neural Networks
(ElmanNN). The evaluation employs two key
performance metrics: Accuracy (ACC) and Mean
Squared Error (MSE). The training dataset,
representing 70% of the data, is used for algorithm
training, and the testing dataset, constituting the
remaining 30%, assesses the algorithms' ability to
generalize to new, unseen data. Results indicate that
SCA_ElmanNN in both training and testing datasets,
achieving high accuracy and minimal MSE, showcasing
its proficiency in classification and prediction precision.
SCA_BP and SCA_ANN also demonstrate robust
performance. Conversely, ElmanNN, while relatively
accurate, exhibits a slightly higher MSE on the testing
data, indicating some variability in its predictions. These
findings offer valuable insights for researchers in
selecting the most appropriate algorithm for specific
classification tasks
Review on Development of Digital Twins for Predicting, Mitigating Faults and Defects in Solar Plants
No full textAbstract – The thought of digital twins has gained substantial attention in recent years due to its potential to transform various industries, including renewable energy. Digital twins involve the creation of virtual models that mirror the behaviour and characteristics of real-world physical systems. In the perspective of solar plants, digital twins have emerged as a promising tool to enhance performance monitoring, predictive maintenance, and overall operational efficiency. Digital twin engineering, characterized by its dynamic data modelling of industrial assets, offers a disruptive technology capable of adapting to real-time changes in the environment and operations. This living model can predict future infrastructure behaviour and proactively identify potential issues within the physical system. The article highlights the essential components of the digital twin ecosystem, such as sensor technologies, the Industrial Internet of Things, simulation, modelling, and machine learning, underscoring their relevance in predictive maintenance applications. This review provides an in-extensive review of the development and application of digital twins for predicting and mitigating faults and defects in solar power plants. It opens with a look at current developments, underlining the rising focus on digital twins for optimizing solar farms. It begins with an overview of existing solutions in the field, highlighting the growing interest in leveraging digital twin technology to enhance solar plant operations. Additionally, the article outlines the implementation stage of a prototype digital twin for a solar power plant
Human Fall Motion Prediction – A Review
No full textAbstract – In predicting human fall motion, focused on enhancing safety and quality of life for the elderly and individuals at risk of falls. By highlighting the critical role of Human Pose Estimation, advancements in human motion forecasting, and fall prediction. It explores the continuous efforts to improve fall detection systems using innovative technologies, such as wearable sensors and IoT devices to implement deep learning models and analyze human poses and gestures. Various methods show promise in accurately predicting human fall motion by capturing complex patterns and relationships in the data. For instance, self-attention mechanisms can revolutionize human motion prediction by effectively capturing these intricate patterns, leading to more accurate predictions. Future research directions should focus on enhancing model accuracy, exploring new techniques for capturing complex patterns, and enabling real-time implementation in wearable devices or smart environments. By addressing these areas, fall detection systems can be significantly improved, benefiting individuals and healthcare systems worldwide