A Systematic Review for Classification and Selection of Deep Learning Methods

Decision Analytics Journal, Journal Year: 2024, Volume and Issue: 12, P. 100489 - 100489

Published: June 5, 2024

The effectiveness of deep learning in completing tasks comprehensively has led to a rapid increase its usage. Deep encompasses numerous diverse methods, each with own distinct characteristics. aim this study is synthesize existing literature order classify and identify an appropriate method for given task. A systematic review was conducted as comprehensive study, utilizing spanning from 2012 2024. findings revealed that plays significant role eight main tasks, including prediction, design, evaluation assessment, decision-making, creating user instructions, classification, identification, models. various such Convolutional Neural Networks (CNN), Recurrent (RNN), Autoencoders (AE), Generative Adversarial (GAN), (DNN), Backpropagation (BP), Feed-Forward (FFNN), different confirmed. These provide researchers understanding selecting effective methods specific tasks.

Language: Английский

---

An investigation of the ensemble machine learning techniques for predicting mechanical properties of printed parts in additive manufacturing

Decision Analytics Journal, Journal Year: 2024, Volume and Issue: 12, P. 100492 - 100492

Published: June 8, 2024

This study investigates the ensemble machine learning models to predict mechanical properties of 3D-printed Polylactic Acid (PLA) specimens. We studied effects five process parameters, including build orientation, infill angle, layer thickness, printing speed, and nozzle temperature, on printed parts tensile strength surface roughness. Machine are developed using experimental data collected from 27 Gradient Boosting Regression, Extreme Adaptive Random Forest Extremely Randomized Tree Regression were during modeling stage roughness parts. research demonstrates effectiveness model in providing accurate predictions with root mean square error (RMSE) 1.03, absolute (MAE) 0.82, percentage (MAPE) 2.20%. Similarly, shows better accuracy predicting having RMSE 0.408, MAE 0.31, MAPE 9.28%. Moreover, comparative confirms that techniques more useful than traditional support vector k-nearest neighbor for The results highlight a novel approach identifying complex correlations dataset, establishing foundation improved product design property optimization through adjustment parameters combination.

Language: Английский

---

An ensemble learning model for forecasting water-pipe leakage

Scientific Reports, Journal Year: 2024, Volume and Issue: 14(1)

Published: May 9, 2024

Based on the benefits of different ensemble methods, such as bagging and boosting, which have been studied adopted extensively in research practice, where boosting focus more reducing variance bias, this paper presented an optimization learning-based model for a large pipe failure dataset water leakage forecasting, something that was not previously considered by others. It is known tuning hyperparameters each base learned inside weight process can produce better-performing ensembles, so it effectively improves accuracy forecasting based pipeline rate. To evaluate proposed model, results are compared with models using root-mean-square error (RMSE), mean square (MSE), absolute (MAE), coefficient determination (R2) technique, technique optimizable higher than other models. The experimental result shows has better prediction accuracy. achieved best rate at 14th iteration, least RMSE = 0.00231 MAE 0.00071513 when building predicts via

Language: Английский

---

An unsupervised machine learning approach for cyber threat detection using geographic profiling and Domain Name System data

Decision Analytics Journal, Journal Year: 2025, Volume and Issue: unknown, P. 100576 - 100576

Published: April 1, 2025

Language: Английский

---

Unveiling the Retention Puzzle for Optimizing Employee Engagement and Loyalty Through Analytics-Driven Performance Management: A Systematic Literature Review

Published: Jan. 28, 2024

Disengagement and turnover of employees are significant costs to organizations worldwide. In many organizations, it isn't easy foster continuous engagement among employees. Analytically-driven performance management aims capture analyze workplace data with advanced analytical techniques develop a sustainable solution. This systematic literature review (SLR) examines analyzes frameworks proposed for optimizing retention through analytics. Among the forty initial papers screened, twenty-four highly relevant sources were selected analyzed. Human resources (HR) related key themes included bias issues, text analysis reviews, personalized HR management, talent assessments, augmenting work Artificial Intelligence (AI), integration challenges. According findings, reliable emphasis was placed on balance human machine perspectives. While analytics algorithms offer insightful information, judgment is needed contextualize this data. If datadriven methods only ones used, complicated personal aspects that influence experience may be overlooked. Consequently, human-machine strategy working together crucial. Furthermore, effective requires both alignment cultural preparedness. Longitudinal evaluations more real-world case studies help close gaps in literature. Analytics human-centric can maximize management.

Language: Английский

---

Analysis and classification of employee attrition and absenteeism in industry: A sequential pattern mining-based methodology

Computers in Industry, Journal Year: 2024, Volume and Issue: 159-160, P. 104106 - 104106

Published: May 27, 2024

Language: Английский

---

Optimization of the algorithms use ensemble and synthetic minority oversampling technique for air quality classification

Indonesian Journal of Electrical Engineering and Computer Science, Journal Year: 2024, Volume and Issue: 33(3), P. 1632 - 1632

Published: Feb. 16, 2024

<p>Rapid economic development, industrialization, and urbanization in Indonesia have caused a large increase air pollution with negative impacts on the environment public health. The aim of this research is to use machine learning techniques categorize quality generate an index (AQI) using dataset that includes six prevalent pollutants. Next steps are preprocessing data extraction, K-nearest neighbors (KNN) classification, support vector (SVM), random forest (RF) models implemented. Furthermore, synthetic minority oversampling technique (SMOTE) incorporated into ensemble process improve results. This uses K-fold cross validation for classification accuracy reduce overfitting. Research findings show application SMOTE brings significant model accuracy, effectively solving problem imbalanced sets. These insights provide direction effective monitoring systems informed decision making management.</p>

Language: Английский

---

An Ensemble Learning Model for Forecasting Water-pipe Leakage

Research Square (Research Square), Journal Year: 2024, Volume and Issue: unknown

Published: Feb. 6, 2024

Based on the benefits of different ensemble methods, such as bagging and boosting, which have been studied adopted extensively in research practice, where boosting focus more reducing variance bias, this paper presented an optimization learning-based model for a large pipe failure dataset water leakage forecasting, something that was not previously considered by others. It is known tuning hyperparameters each base learned inside weight process can produce better-performing ensembles, so it effectively improves accuracy forecasting based pipeline rate. To evaluate proposed model, results are compared with models using root-mean-square error (RMSE), mean square (MSE), absolute (MAE), coefficient determination (R2) technique, technique optimizable higher than other models. The experimental result shows has better prediction accuracy. achieved best rate at 14th iteration, least RMSE = 0.00231 MAE 0.00071513 when building predicts via

Language: Английский

---

Exploring Employee Retention among Generation Z Engineers in the Philippines Using Machine Learning Techniques

Sustainability, Journal Year: 2024, Volume and Issue: 16(12), P. 5207 - 5207

Published: June 19, 2024

Generation Z represents a significant portion of the current workforce and is poised to become dominant in engineering field. As new generation arises, employee retention becomes crucial topic Philippines. Hence, this study explored factors influencing among engineers Philippines using machine learning feature selection (filter method’s permutation, wrapper backward elimination, embedded Least Absolute Shrinkage Selection Operator) classifiers (support vector random forest). A total 412 participants were gathered through purposive sampling technique. The results showed that six out seven investigated features found be impacting engineers’ intention remain company. These supervisor support, company attachment, job satisfaction, contribution, emotional shared value, organized descending order importance. further explained by fifteen subfeatures representing each feature. Only one feature, servant leadership, was deemed insignificant. findings extracted from optimal combination algorithms. Particularly, selection’s elimination brought 85.66% accuracy, forest classifier enhanced accuracy value 90.10%. In addition, model’s precision, recall, F1-score values 89.50%, 90.10%, 88.90%, respectively. This research also provided practical insights for executives, organizational leaders, human resources department seeking enhance strategies. implications based on retention, ultimately contributing long-term success competitiveness organizations.

Language: Английский

---

Artificial Neural Networks (ANNs) and Machine Learning (ML) Modeling Employee Behavior with Management Towards the Economic Advancement of Workers

Sustainability, Journal Year: 2024, Volume and Issue: 16(21), P. 9516 - 9516

Published: Nov. 1, 2024

The role of employee behavior in organizations and their interaction with management is crucial advancing the economic progress workers. This study examines impact practices on organizational performance progress, using advanced artificial intelligence techniques to explore complex relationships provide evidence-based strategies for sustainable workforce development. research analyzes critical aspects such as job satisfaction, motivation, participation, communication uncover underlying mechanisms that contribute It recognizes dynamic relationship between employees management, confirming central effective leadership, communication, teamwork achieving positive results. emphasizes harmonious cooperation necessary create a favorable work environment contributes development utilizes an neural network (ANN) better understand interdependencies different parameters effects within framework this ongoing project. results existing body knowledge by providing practical implications seeking optimize employee–employer increase overall productivity. By understanding dynamics practices, can supportive maximizes potential growth. findings demonstrate accuracy over 70%, indicating enhancing satisfaction significantly improve productivity, performance.

Language: Английский

---