Ho Seung Song
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Ho Seung Song has a total of 36 co-authors. Here are the top co-authors by number of shared publications:
- Mitra Ghergherehchi
Affiliation Department of Electrical and Computer Engineering, Sungkyunkwan University, Suwon, South Korea
6 shared publications - Jong-Seo Chai
Department of Electrical and Computer Engineering, Sungkyunkwan University, Suwon, Korea
3 shared publications Jongchul Lee
3 shared publicationsDonghyup Ha
2 shared publications- Seung Hyun Lee
Sungkyunkwan University
2 shared publications - Berat Can Karatas
College of Information & Communication Engineering, Sungkyunkwan University, Suwon-si, Republic of Korea
1 shared publication
Latest Publications
- Open Access
- 23 February 2026
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 21 (2) , e0342062
Trustworthy AI for medical decisions: Adversarially robust and fair machine learning prediction for Parkinson’s disease
Parkinson’s disease (PD) is a neurodegenerative disorder characterized by motor and non-motor symptoms, including tremor, rigidity, and postural instability. Machine learning (ML) models have shown promise for the diagnosis of PD; however, many existing approaches do not explicitly address fairness and robustness. As a result, these models can lead to biased outcomes across demographic groups and vulnerability to adversarial attacks. In this study, we used the Parkinson’s Progression Markers Initiative (PPMI) cohort, which includes clinical and demographic information from 1,084 participants spanning diverse age, sex, and racial groups. Our study addresses the key challenge of developing robust and equitable ML models to diagnose the progression of PD. We evaluated the performance of two fairness-optimized classifiers, namely, Random Forest (RF) and Decision Tree (DT). To evaluate model vulnerability, we applied adversarial techniques, specifically label leakage and data poisoning attacks, which simulate intentional or erroneous data alterations that can amplify biases and degrade accuracy. These adversarial manipulations substantially degraded model performance; specifically, DT accuracy declined by more than 10% between sensitive groups. The accuracy of the RF model decreased by 20%. Moreover, under attack, fairness metrics such as Statistical Parity Difference (SPD), which looks at differences in the chances of getting a positive prediction across demographic groups, and Equal Opportunity Difference (EOD) for differences in true positive rates between groups, both showed a decline. This pattern suggests that adversarial perturbations increased bias and widened performance disparities across demographic groups. Our results demonstrated that adversarial attacks increased the incidence of false positives and false negatives, thereby lowering the accuracy and fairness of the PD diagnostic predictions. These findings underscore the urgent need for robust and fairness-aware defenses in medical AI to mitigate racial, age, and gender disparities and ensure a reliable clinical decision-making process.
- 23 February 2026
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 21 (2) , e0342062
Open AccessTrustworthy AI for medical decisions: Adversarially robust and fair machine learning prediction for Parkinson’s disease
Parkinson’s disease (PD) is a neurodegenerative disorder characterized by motor and non-motor symptoms, including tremor, rigidity, and postural instability. Machine learning (ML) models have shown promise for the diagnosis of PD; however, many existing approaches do not explicitly address fairness and robustness. As a result, these models can lead to biased outcomes across demographic groups and vulnerability to adversarial attacks. In this study, we used the Parkinson’s Progression Markers Initiative (PPMI) cohort, which includes clinical and demographic information from 1,084 participants spanning diverse age, sex, and racial groups. Our study addresses the key challenge of developing robust and equitable ML models to diagnose the progression of PD. We evaluated the performance of two fairness-optimized classifiers, namely, Random Forest (RF) and Decision Tree (DT). To evaluate model vulnerability, we applied adversarial techniques, specifically label leakage and data poisoning attacks, which simulate intentional or erroneous data alterations that can amplify biases and degrade accuracy. These adversarial manipulations substantially degraded model performance; specifically, DT accuracy declined by more than 10% between sensitive groups. The accuracy of the RF model decreased by 20%. Moreover, under attack, fairness metrics such as Statistical Parity Difference (SPD), which looks at differences in the chances of getting a positive prediction across demographic groups, and Equal Opportunity Difference (EOD) for differences in true positive rates between groups, both showed a decline. This pattern suggests that adversarial perturbations increased bias and widened performance disparities across demographic groups. Our results demonstrated that adversarial attacks increased the incidence of false positives and false negatives, thereby lowering the accuracy and fairness of the PD diagnostic predictions. These findings underscore the urgent need for robust and fairness-aware defenses in medical AI to mitigate racial, age, and gender disparities and ensure a reliable clinical decision-making process.
- Open Access
- 24 January 2025
- Published by Elsevier in Engineering Science and Technology, an International Journal
- Vol. 62
Memory-driven deep-reinforcement learning for autonomous robot navigation in partially observable environments
- 24 January 2025
- Published by Elsevier in Engineering Science and Technology, an International Journal
- Vol. 62
Open AccessMemory-driven deep-reinforcement learning for autonomous robot navigation in partially observable environments
- Open Access
- 9 September 2023
- Published by Elsevier in Nuclear Engineering and Technology
- Vol. 55 (12) , 4570-4575
Real-time 14N NQR-based sodium nitrite analysis in a noisy field
- 9 September 2023
- Published by Elsevier in Nuclear Engineering and Technology
- Vol. 55 (12) , 4570-4575
Open AccessReal-time 14N NQR-based sodium nitrite analysis in a noisy field
- Open Access
- 26 June 2023
- Published by Elsevier in Nuclear Engineering and Technology
- Vol. 55 (8) , 3114-3120
Development of magnetic field measurement system for AMS cyclotron
- 26 June 2023
- Published by Elsevier in Nuclear Engineering and Technology
- Vol. 55 (8) , 3114-3120
Open AccessDevelopment of magnetic field measurement system for AMS cyclotron
- 25 October 2021
- Published by Springer Nature in Journal of the Korean Physical Society
- Vol. 79 (9) , 857-863
Improvement of the RF cavity for the SKKUCY-10 cyclotron
- 25 October 2021
- Published by Springer Nature in Journal of the Korean Physical Society
- Vol. 79 (9) , 857-863
Improvement of the RF cavity for the SKKUCY-10 cyclotron
- 1 March 2017
- Published by AIP Publishing in Review of Scientific Instruments
- Vol. 88 (3) , 034706
Development of an 83.2 MHz, 3.2 kW solid-state RF amplifier using Wilkinson power divider and combiner for a 10 MeV cyclotron
We design a stripline-type Wilkinson power divider and combiner for a 3.2 kW solid-state radio frequency (RF) amplifier module and optimize this setup. A Teflon-based printed circuit board is used in the power combiner to transmit high RF power efficiently in the limited space. The reflection coefficient (S11) and insertion loss (S21) related to impedance matching are characterized to determine the optimization process. The resulting two-way divider reflection coefficient and insertion loss were −48.00 dB and −3.22 dB, respectively. The two-way power combiner reflection coefficient and insertion loss were −20 dB and −3.3 dB, respectively. Moreover, the 3.2 kW solid-state RF power test results demonstrate that the proposed power divider and combiner exhibit a maximum efficiency value of 71.3% (combiner loss 5%) at 48 V supply voltage.
- 1 March 2017
- Published by AIP Publishing in Review of Scientific Instruments
- Vol. 88 (3) , 034706
Development of an 83.2 MHz, 3.2 kW solid-state RF amplifier using Wilkinson power divider and combiner for a 10 MeV cyclotron
We design a stripline-type Wilkinson power divider and combiner for a 3.2 kW solid-state radio frequency (RF) amplifier module and optimize this setup. A Teflon-based printed circuit board is used in the power combiner to transmit high RF power efficiently in the limited space. The reflection coefficient (S11) and insertion loss (S21) related to impedance matching are characterized to determine the optimization process. The resulting two-way divider reflection coefficient and insertion loss were −48.00 dB and −3.22 dB, respectively. The two-way power combiner reflection coefficient and insertion loss were −20 dB and −3.3 dB, respectively. Moreover, the 3.2 kW solid-state RF power test results demonstrate that the proposed power divider and combiner exhibit a maximum efficiency value of 71.3% (combiner loss 5%) at 48 V supply voltage.