Trung Nguyen Thanh
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Trung Nguyen Thanh has a total of 8 co-authors. Here are the top co-authors by number of shared publications:
- Tai Le Quy
University of Koblenz, Koblenz, Germany
4 shared publications - Ha-Bang Ban
Hanoi University of Science and Technology, Hanoi, Vietnam
2 shared publications - Huyen Giang Thi Thu
Graduate University of Science and Technology, Vietnam Academy of Science and Technology, Hanoi, Vietnam
2 shared publications - Dung Nguyen Tuan1 shared publication
Duy Tran Cong
1 shared publication- Frank Hopfgartner
University of Koblenz
1 shared publication
Latest Publications
- 2 August 2026
- Published by Springer Nature
- p. 223-233
Constraint-Reduced MILP with Local Outlier Factor Modeling for Plausible Counterfactual Explanations in Credit Approval
- 2 August 2026
- Published by Springer Nature
- p. 223-233
Constraint-Reduced MILP with Local Outlier Factor Modeling for Plausible Counterfactual Explanations in Credit Approval
- 25 June 2026
- Published by Springer Nature
- p. 190-202
FairFinGAN: Fairness-Aware Synthetic Financial Data Generation
- 25 June 2026
- Published by Springer Nature
- p. 190-202
FairFinGAN: Fairness-Aware Synthetic Financial Data Generation
- Open Access
Constraint-Reduced MILP with Local Outlier Factor Modeling for Plausible Counterfactual Explanations in Credit Approval
Counterfactual explanation (CE) is a widely used post-hoc method that provides individuals with actionable changes to alter an unfavorable prediction from a machine learning model. Plausible CE methods improve realism by considering data distribution characteristics, but their optimization models introduce a large number of constraints, leading to high computational cost. In this work, we revisit the DACE framework and propose a refined Mixed-Integer Linear Programming (MILP) formulation that significantly reduces the number of constraints in the local outlier factor (LOF) objective component. We also apply the method to a linear SVM classifier with standard scaler. The experimental results show that our approach achieves faster solving times while maintaining explanation quality. These results demonstrate the promise of more efficient LOF modeling in counterfactual explanation and data science applications.
Open AccessConstraint-Reduced MILP with Local Outlier Factor Modeling for Plausible Counterfactual Explanations in Credit Approval
Counterfactual explanation (CE) is a widely used post-hoc method that provides individuals with actionable changes to alter an unfavorable prediction from a machine learning model. Plausible CE methods improve realism by considering data distribution characteristics, but their optimization models introduce a large number of constraints, leading to high computational cost. In this work, we revisit the DACE framework and propose a refined Mixed-Integer Linear Programming (MILP) formulation that significantly reduces the number of constraints in the local outlier factor (LOF) objective component. We also apply the method to a linear SVM classifier with standard scaler. The experimental results show that our approach achieves faster solving times while maintaining explanation quality. These results demonstrate the promise of more efficient LOF modeling in counterfactual explanation and data science applications.
- Open Access
- 3 April 2025
- Published by Learning Gate in Edelweiss Applied Science and Technology
- Vol. 9 (4) , 1962-1974
Machine learning and statistical modeling in firm value prediction
Determining the best approach model—machine learning or statistical model—to solve the analysis challenges of firm value prediction has garnered attention from many scholars. This article used both traditional statistical models and machine learning to predict the firm value of 435 non-financial companies that are listed on the Vietnam Security Exchange during 2014-2021. Based on the empirical results, the machine learning models provided evidence that they forecast firm value better than traditional models and identify the number of firm value determinants. The paper applied six machine learning models to find the best-performing one, including the multiple regression model (LM), Lasso, generalized additive model (GAM), random forests (RF), gradient boosting regression trees (GBM), and neural networks (NNET). The findings indicated that the RF is the best-performing model, selecting firm size, ROA, tangibility, GDP, quality, financial leverage, and inflation as reliable predictors of market firm value. The results suggest several recommendations for internal managers, investors, and creditors when choosing the appropriate model to forecast firm value for making financial decisions.
- 3 April 2025
- Published by Learning Gate in Edelweiss Applied Science and Technology
- Vol. 9 (4) , 1962-1974
Open AccessMachine learning and statistical modeling in firm value prediction
Determining the best approach model—machine learning or statistical model—to solve the analysis challenges of firm value prediction has garnered attention from many scholars. This article used both traditional statistical models and machine learning to predict the firm value of 435 non-financial companies that are listed on the Vietnam Security Exchange during 2014-2021. Based on the empirical results, the machine learning models provided evidence that they forecast firm value better than traditional models and identify the number of firm value determinants. The paper applied six machine learning models to find the best-performing one, including the multiple regression model (LM), Lasso, generalized additive model (GAM), random forests (RF), gradient boosting regression trees (GBM), and neural networks (NNET). The findings indicated that the RF is the best-performing model, selecting firm size, ROA, tangibility, GDP, quality, financial leverage, and inflation as reliable predictors of market firm value. The results suggest several recommendations for internal managers, investors, and creditors when choosing the appropriate model to forecast firm value for making financial decisions.
- 28 April 2022
- Published by Springer Nature in SN Computer Science
- Vol. 3 (3) , 1-14
Dynamic Sliding Window and Neighborhood LSTM-Based Model for Stock Price Prediction
- 28 April 2022
- Published by Springer Nature in SN Computer Science
- Vol. 3 (3) , 1-14
Dynamic Sliding Window and Neighborhood LSTM-Based Model for Stock Price Prediction