DLSTM-SCM– A Dynamic LSTM-Based Framework for Smart Supply Chain Management DOI

Seyf Eddine Hasnaoui,

Mohammed Amine Boudouaia, Samir Ouchani

et al.

Published: Dec. 14, 2023

In the retail industry, SCM holds significant importance as it ensures efficient movement of goods from suppliers to customers.In this intricate and fast-paced environment, availability accurate information data is crucial.The purpose paper develop a framework that enhances forecasting accuracy efficiency in supply chain operations within industry.By analyzing latest research advancements field, seeks contribute valuable insights into potential deep learning for management.The ultimate goal provide retailers with reliable tool empowers them make informed decisions based on predictions, thereby optimizing their better meeting customer demands dynamic landscape.DLSTM-SCM, developed paper, updates dynamically deployed LSTM models predict upcoming day's sales using historical addition statistical features like lagging shifting enhance precision.The efficacy DLSTM-SCM demonstrated through its performance real benchmarks, where yielded improvements compared existing methods.

Language: Английский

A systematic review of hyperparameter optimization techniques in Convolutional Neural Networks DOI Creative Commons
Mohaimenul Azam Khan Raiaan, Sadman Sakib, Nur Mohammad Fahad

et al.

Decision Analytics Journal, Journal Year: 2024, Volume and Issue: 11, P. 100470 - 100470

Published: April 24, 2024

Convolutional Neural Network (CNN) is a prevalent topic in deep learning (DL) research for their architectural advantages. CNN relies heavily on hyperparameter configurations, and manually tuning these hyperparameters can be time-consuming researchers, therefore we need efficient optimization techniques. In this systematic review, explore range of well used algorithms, including metaheuristic, statistical, sequential, numerical approaches, to fine-tune hyperparameters. Our offers an exhaustive categorization (HPO) algorithms investigates the fundamental concepts CNN, explaining role variants. Furthermore, literature review HPO employing above mentioned undertaken. A comparative analysis conducted based strategies, error evaluation accuracy results across various datasets assess efficacy methods. addition addressing current challenges HPO, our illuminates unresolved issues field. By providing insightful evaluations merits demerits objective assist researchers determining suitable method particular problem dataset. highlighting future directions synthesizing diversified knowledge, survey contributes significantly ongoing development optimization.

Language: Английский

Citations

40
Advances in artificial intelligence for drug delivery and development: A comprehensive review DOI
Amol D. Gholap, Md Jasim Uddin, Md. Faiyazuddin

et al.

Computers in Biology and Medicine, Journal Year: 2024, Volume and Issue: 178, P. 108702 - 108702

Published: June 7, 2024

Language: Английский

Citations

37
Hyperparameter Optimization of Ensemble Models for Spam Email Detection DOI Creative Commons
Temidayo Oluwatosin Omotehinwa, David Opeoluwa Oyewola

Applied Sciences, Journal Year: 2023, Volume and Issue: 13(3), P. 1971 - 1971

Published: Feb. 3, 2023

Unsolicited emails, popularly referred to as spam, have remained one of the biggest threats cybersecurity globally. More than half emails sent in 2021 were resulting huge financial losses. The tenacity and perpetual presence adversary, spammer, has necessitated need for improved efforts at filtering spam. This study, therefore, developed baseline models random forest extreme gradient boost (XGBoost) ensemble algorithms detection classification spam using Enron1 dataset. then optimized grid-search cross-validation technique search hyperparameter space optimal values. performance (un-tuned) tuned both evaluated compared. impact tuning on was also examined. findings experimental study revealed that when compared with models. RF XGBoost achieved an accuracy 97.78% 98.09%, a sensitivity 98.44% 98.84%, F1 score 97.85% 98.16%, respectively. model outperformed model. is effective efficient email detection.

Language: Английский

Citations

32
A Light Gradient-Boosting Machine algorithm with Tree-Structured Parzen Estimator for breast cancer diagnosis DOI Creative Commons
Temidayo Oluwatosin Omotehinwa, David Opeoluwa Oyewola, Emmanuel Gbenga Dada

et al.

Healthcare Analytics, Journal Year: 2023, Volume and Issue: 4, P. 100218 - 100218

Published: June 24, 2023

Breast cancer is a common and potentially life-threatening disease. Early accurate diagnosis of breast crucial for effective treatment improved patient outcomes. This study proposed using the Light Gradient-Boosting Machine (LightGBM) algorithm, Borderline- Synthetic Minority Oversampling Technique (SMOTE), Tree-Structured Parzen Estimator (TPE) hyperparameter tuning to enhance effectiveness Learning (ML) model diagnosing cancer. A 10-fold cross-validated TPE optimized Borderline-SMOTE LightGBM classifier was modelled on Wisconsin Diagnostic Cancer (WDBC) Dataset evaluated its performance compared baseline model. The TPE-optimized exhibited significant improvement in over model, achieving an average accuracy 99.12%, specificity 100%, precision recall 97.62%, F1-score 98.80%, Mathews Correlation Coefficient 98.12%. Compared previous studies, performed exceptionally well. demonstrates data augmentation optimization techniques improve ML models diagnosis, which has implications medical field where efficient critical.

Language: Английский

Citations

30
Optimizing the light gradient-boosting machine algorithm for an efficient early detection of coronary heart disease DOI Creative Commons
Temidayo Oluwatosin Omotehinwa, David Opeoluwa Oyewola,

Ervin Gubin Moung

et al.

Informatics and Health, Journal Year: 2024, Volume and Issue: 1(2), P. 70 - 81

Published: July 2, 2024

Coronary heart disease (CHD) remains a prominent cause of mortality globally, necessitating early and accurate detection methods. Traditional diagnostic approaches can be invasive, costly, time-consuming, the need for more efficient alternatives. This aimed to optimize Light Gradient-Boosting Machine (LightGBM) algorithm enhance its performance accuracy in CHD, providing reliable, cost-effective, non-invasive tool. The Framingham Heart Study (FHS) dataset publicly available on Kaggle was used this study. Multiple Imputations by Chained Equations (MICE) were applied separately training testing sets handle missing data. Borderline-SMOTE (Synthetic Minority Over-sampling Technique) set balance dataset. LightGBM selected efficiency classification tasks, Bayesian Optimization with Tree-structured Parzen Estimator (TPE) employed fine-tune hyperparameters. optimized model trained evaluated using metrics such as accuracy, precision, AUC-ROC test set, cross-validation ensure robustness generalizability. showed significant improvement CHD detection. baseline dropped values had an 0.8333, sensitivity 0.1081, precision 0.3429, F1 score 0.1644, AUC 0.6875. With MICE imputation, improved 0.9399, 0.6693, 0.9043, 0.7692, 0.9457. combined approach Borderline-SMOTE, TPE achieved 0.9882, 0.9370, 0.9835, 0.9597, 0.9963, indicating highly effective robust model. demonstrated outstanding study's strengths include comprehensive addressing data class imbalance fine-tuning hyperparameters through Optimization. However, there is other datasets generalizability well-established. study provides strong framework detection, improving clinical practice allowing precise dependable diagnostics interventions.

Language: Английский

Citations

9
Predicting Demand in Supply Chain Networks With Quantum Machine Learning Approach DOI
Sunil Kumar Sehrawat, Pushan Kumar Dutta,

A. Bhatia

et al.

Advances in logistics, operations, and management science book series, Journal Year: 2024, Volume and Issue: unknown, P. 33 - 47

Published: June 30, 2024

This chapter explores the application of quantum machine learning (QML) techniques for demand prediction in supply chain networks. Traditional forecasting methods often struggle to capture intricate dynamics and uncertainties present modern chains. By leveraging computational power probabilistic nature computing, coupled with flexibility adaptability algorithms, organizations can enhance accuracy efficiency their processes. provides an overview QML methodologies tailored specifically networks, highlighting advantages over classical approaches. Through case studies practical examples, demonstrates how enable make more informed decisions, optimize inventory levels, improve overall performance.

Language: Английский

Citations

4
Optimizing sentiment analysis of Nigerian 2023 presidential election using two-stage residual long short term memory DOI Creative Commons
David Opeoluwa Oyewola, Lawal Abdullahi Oladimeji,

Sowore Olatunji Julius

et al.

Heliyon, Journal Year: 2023, Volume and Issue: 9(4), P. e14836 - e14836

Published: March 30, 2023

Sentiment analysis is the process of recognizing positive or negative attitudes in text. This technique makes use computational linguistics, text analysis, and natural language processing. The 2023 presidential election Nigeria a significant event for country, as it will determine leader nation next four years. As such, important to understand sentiment public towards different candidates. In this research, we aimed three main candidates Nigeria, Atiku, Tinubu, Obi, by conducting on tweets related We used long short-term memory (LSTM), peephole short term (PLSTM), two-stage residual (TSRLSTM) models classify positive, neutral, negative. Our dataset consisted large number that were preprocessed remove noise irrelevant information. Results showed TSRLSTM performed excellently well classifying identifying each candidate individually. findings provide valuable insights into public's opinion their campaign strategies, which can be useful researchers, political analysts, decision-makers. study highlights importance understanding its potential applications field science.

Language: Английский

Citations

9
Hyperparameter optimization for deep neural network models: a comprehensive study on methods and techniques DOI
Sunita Roy, Ranjan Mehera, Rajat Kumar Pal

et al.

Innovations in Systems and Software Engineering, Journal Year: 2023, Volume and Issue: unknown

Published: Oct. 7, 2023

Language: Английский

Citations

7
Optimizing Supply Chain Management Through BO-CNN-LSTM for Demand Forecasting and Inventory Management DOI Open Access
Rong Liu, Vinay Vakharia

Journal of Organizational and End User Computing, Journal Year: 2024, Volume and Issue: 36(1), P. 1 - 25

Published: Jan. 7, 2024

This project addresses demand forecasting and inventory optimization in supply chain management. Traditional methods have limitations with complex patterns large-scale data. Deep learning techniques are employed to enhance accuracy efficiency. The utilizes BO-CNN-LSTM, leveraging Bayesian for hyperparameter tuning, Convolutional Neural Networks (CNNs) spatiotemporal feature extraction, Long Short-Term Memory (LSTMs) modeling sequential Experimental results validate the effectiveness of approach, outperforming traditional methods. Practical implementation management improves operational efficiency cost control.

Language: Английский

Citations

2
Deep Learning for Predicting Attrition Rate in Open and Distance Learning (ODL) Institutions DOI Creative Commons
Juliana Ngozi Ndunagu, David Opeoluwa Oyewola, Farida Shehu Garki

et al.

Computers, Journal Year: 2024, Volume and Issue: 13(9), P. 229 - 229

Published: Sept. 11, 2024

Student enrollment is a vital aspect of educational institutions, encompassing active, registered and graduate students. All the same, some students fail to engage with their studies after admission drop out along line; this known as attrition. The student attrition rate acknowledged most complicated significant problem facing systems caused by institutional non-institutional challenges. In study, researchers utilized dataset obtained from National Open University Nigeria (NOUN) 2012 2022, which included comprehensive information about enrolled in various programs at university who were inactive had dropped out. used deep learning techniques, such Long Short-Term Memory (LSTM) model compared performance One-Dimensional Convolutional Neural Network (1DCNN) model. results study revealed that LSTM achieved overall accuracy 57.29% on training data, while 1DCNN exhibited lower 49.91% data. indicated superior correct classification

Language: Английский

Citations

2