Hao Luo
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
Hao Luo has a total of 24 co-authors. Here are the top co-authors by number of shared publications:
- Hung-Yu Wei
National Taiwan University
4 shared publications - Chi-Hung Liu
School of Medicine, College of Medicine, Chang Gung University, Taoyuan City 33302, Taiwan
2 shared publications - Ching-Po Lin
Institute of Neuroscience, National Yang Ming Chiao Tung University, Taipei, Taiwan
2 shared publications - Kuan-Fu Chen
Department of Emergency Medicine, Keelung Chang Gung Memorial Hospital, Keelung, Taiwan
2 shared publications - Po-Yuan Su
Department of Electrical Engineering National Taiwan University Taipei Taiwan
2 shared publications - Tsong-Hai Lee
Stroke Center and Department of Neurology, Linkou Chang Gung Memorial Hospital and College of Medicine, Chang Gung University, Taoyuan, Taiwan (T.-H.L.).
2 shared publications
Latest Publications
- 7 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Communications
- Vol. 73 (12) , 15909-15925
Digital Twin Aided Massive MIMO CSI Feedback: Exploring the Impact of Twinning Fidelity
Deep learning (DL) techniques have demonstrated strong performance in compressing and reconstructing channel state information (CSI) while reducing feedback overhead in massive MIMO systems. A key challenge, however, is their reliance on extensive site-specific training data, whose real-world collection incurs significant overhead and limits scalability across deployment sites. To address this, we propose leveraging site-specific digital twins to assist the training of DL-based CSI compression models. The digital twin integrates an electromagnetic (EM) 3D model of the environment, a hardware model, and ray tracing to produce site-specific synthetic CSI data, allowing DL models to be trained without the need for extensive real-world measurements. We further develop a fidelity analysis framework that decomposes digital twin quality into four key aspects: 3D geometry, material properties, ray tracing, and hardware modeling. We explore how these factors influence the reliability of the data and model performance. To enhance the adaptability to real-world environments, we propose a refinement strategy that incorporates a limited amount of real-world data to fine-tune the DL model pre-trained on the digital twin dataset. Evaluation results show that models trained on site-specific digital twins outperform those trained on generic datasets, with the proposed refinement method effectively enhancing performance using limited real-world data. The simulations also highlight the importance of digital twin fidelity, especially in 3D geometry, ray tracing, and hardware modeling, for improving CSI reconstruction quality. This analysis framework offers valuable insights into the critical fidelity aspects, and facilitates more efficient digital twin development and deployment strategies for various wireless communication tasks.
- 7 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Communications
- Vol. 73 (12) , 15909-15925
Digital Twin Aided Massive MIMO CSI Feedback: Exploring the Impact of Twinning Fidelity
Deep learning (DL) techniques have demonstrated strong performance in compressing and reconstructing channel state information (CSI) while reducing feedback overhead in massive MIMO systems. A key challenge, however, is their reliance on extensive site-specific training data, whose real-world collection incurs significant overhead and limits scalability across deployment sites. To address this, we propose leveraging site-specific digital twins to assist the training of DL-based CSI compression models. The digital twin integrates an electromagnetic (EM) 3D model of the environment, a hardware model, and ray tracing to produce site-specific synthetic CSI data, allowing DL models to be trained without the need for extensive real-world measurements. We further develop a fidelity analysis framework that decomposes digital twin quality into four key aspects: 3D geometry, material properties, ray tracing, and hardware modeling. We explore how these factors influence the reliability of the data and model performance. To enhance the adaptability to real-world environments, we propose a refinement strategy that incorporates a limited amount of real-world data to fine-tune the DL model pre-trained on the digital twin dataset. Evaluation results show that models trained on site-specific digital twins outperform those trained on generic datasets, with the proposed refinement method effectively enhancing performance using limited real-world data. The simulations also highlight the importance of digital twin fidelity, especially in 3D geometry, ray tracing, and hardware modeling, for improving CSI reconstruction quality. This analysis framework offers valuable insights into the critical fidelity aspects, and facilitates more efficient digital twin development and deployment strategies for various wireless communication tasks.
- Top Cited Papers
- 16 February 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 10 (16) , 14307-14331
Management and Orchestration of Edge Computing for IoT: A Comprehensive Survey
With the development of telecommunication technologies and the proliferation of network applications in the past decades, the traditional cloud network architecture becomes unable to accommodate such demands due to the heavy burden on the backhaul links and long latency. Therefore, edge computing, which brings network functions close to end-users by providing caching, computing and communication resources at network edges, turns into a promising paradigm. Benefit from its nature, edge computing enables emerging scenarios and use cases, such as augmented reality (AR) and Internet of Things (IowT). However, it also creates complexities to efficiently orchestrate heterogeneous services and manage distributed resources in the edge network. In this survey, we make a comprehensive review of the research efforts on service orchestration and resource management for edge computing. We first give an overview of edge computing, including architectures, advantages, enabling technologies and standardization. Next, a comprehensive survey of state-of-the-art techniques in the management and orchestration of edge computing is presented. Subsequently, the state-of-the-art research on the infrastructure of edge computing is discussed in various aspects. Finally, open research challenges and future directions are presented as well.
- 16 February 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 10 (16) , 14307-14331
Top Cited PapersManagement and Orchestration of Edge Computing for IoT: A Comprehensive Survey
With the development of telecommunication technologies and the proliferation of network applications in the past decades, the traditional cloud network architecture becomes unable to accommodate such demands due to the heavy burden on the backhaul links and long latency. Therefore, edge computing, which brings network functions close to end-users by providing caching, computing and communication resources at network edges, turns into a promising paradigm. Benefit from its nature, edge computing enables emerging scenarios and use cases, such as augmented reality (AR) and Internet of Things (IowT). However, it also creates complexities to efficiently orchestrate heterogeneous services and manage distributed resources in the edge network. In this survey, we make a comprehensive review of the research efforts on service orchestration and resource management for edge computing. We first give an overview of edge computing, including architectures, advantages, enabling technologies and standardization. Next, a comprehensive survey of state-of-the-art techniques in the management and orchestration of edge computing is presented. Subsequently, the state-of-the-art research on the infrastructure of edge computing is discussed in various aspects. Finally, open research challenges and future directions are presented as well.
- 27 September 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Network and Service Management
- Vol. 20 (1) , 385-399
Resource Orchestration at the Edge: Intelligent Management of mmWave RAN and Gaming Application QoE Enhancement
Millimeter wave (mmWave) is a crucial component in 5G and beyond 5G communications. However, the dense deployment of mmWave transceivers would impose a heavy burden on the management of the radio access network (RAN). This challenge increases the need for leveraging intelligent network management techniques. Thanks to edge computing, machine learning (ML) based network management algorithms and other delay-sensitive user applications can operate at the network edge. But, due to the limited resources on edge servers, developing an orchestration scheme for intelligent network management and user applications is necessary. In this paper, we provide an edge-centric resource management framework for intelligent RAN management and applications with the awareness of the users’ quality of experiences (QoE). Specifically, we consider the scenario of a mmWave communication system equipped with an ML-based mmWave beam tracking algorithm. The users under this system request mobile edge gaming services. We formulate a game QoE aware orchestration problem as a non-linear integer programming and prove its NP-hardness. To reduce the complexity, we decompose the original problem into two subproblems, the service placement problem for mobile edge gaming and the configuration selection and placement problem for mmWave beam tracking. Then, we solve the two subproblems consecutively with heuristic approaches. Simulation results demonstrate the effectiveness of the proposed orchestration scheme.
- 27 September 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Network and Service Management
- Vol. 20 (1) , 385-399
Resource Orchestration at the Edge: Intelligent Management of mmWave RAN and Gaming Application QoE Enhancement
Millimeter wave (mmWave) is a crucial component in 5G and beyond 5G communications. However, the dense deployment of mmWave transceivers would impose a heavy burden on the management of the radio access network (RAN). This challenge increases the need for leveraging intelligent network management techniques. Thanks to edge computing, machine learning (ML) based network management algorithms and other delay-sensitive user applications can operate at the network edge. But, due to the limited resources on edge servers, developing an orchestration scheme for intelligent network management and user applications is necessary. In this paper, we provide an edge-centric resource management framework for intelligent RAN management and applications with the awareness of the users’ quality of experiences (QoE). Specifically, we consider the scenario of a mmWave communication system equipped with an ML-based mmWave beam tracking algorithm. The users under this system request mobile edge gaming services. We formulate a game QoE aware orchestration problem as a non-linear integer programming and prove its NP-hardness. To reduce the complexity, we decompose the original problem into two subproblems, the service placement problem for mobile edge gaming and the configuration selection and placement problem for mmWave beam tracking. Then, we solve the two subproblems consecutively with heuristic approaches. Simulation results demonstrate the effectiveness of the proposed orchestration scheme.
- Open Access
- 25 March 2022
- Published by JMIR Publications Inc. in JMIR Medical Informatics
- Vol. 10 (3) , e32508
Machine Learning Models for Predicting Influential Factors of Early Outcomes in Acute Ischemic Stroke: Registry-Based Study
Background: Timely and accurate outcome prediction plays a vital role in guiding clinical decisions on acute ischemic stroke. Early condition deterioration and severity after the acute stage are determinants for long-term outcomes. Therefore, predicting early outcomes is crucial in acute stroke management. However, interpreting the predictions and transforming them into clinically explainable concepts are as important as the predictions themselves. Objective: This work focused on machine learning model analysis in predicting the early outcomes of ischemic stroke and used model explanation skills in interpreting the results. Methods: Acute ischemic stroke patients registered on the Stroke Registry of the Chang Gung Healthcare System (SRICHS) in 2009 were enrolled for machine learning predictions of the two primary outcomes: modified Rankin Scale (mRS) at hospital discharge and in-hospital deterioration. We compared 4 machine learning models, namely support vector machine (SVM), random forest (RF), light gradient boosting machine (LGBM), and deep neural network (DNN), with the area under the curve (AUC) of the receiver operating characteristic curve. Further, 3 resampling methods, random under sampling (RUS), random over sampling, and the synthetic minority over-sampling technique, dealt with the imbalanced data. The models were explained based on the ranking of feature importance and the SHapley Additive exPlanations (SHAP). Results: RF performed well in both outcomes (discharge mRS: mean AUC 0.829, SD 0.018; in-hospital deterioration: mean AUC 0.710, SD 0.023 on original data and 0.728, SD 0.036 on resampled data with RUS for imbalanced data). In addition, DNN outperformed other models in predicting in-hospital deterioration on data without resampling (mean AUC 0.732, SD 0.064). In general, resampling contributed to the limited improvement of model performance in predicting in-hospital deterioration using imbalanced data. The features obtained from the National Institutes of Health Stroke Scale (NIHSS), white blood cell differential counts, and age were the key features for predicting discharge mRS. In contrast, the NIHSS total score, initial blood pressure, having diabetes mellitus, and features from hemograms were the most important features in predicting in-hospital deterioration. The SHAP summary described the impacts of the feature values on each outcome prediction. Conclusions: Machine learning models are feasible in predicting early stroke outcomes. An enriched feature bank could improve model performance. Initial neurological levels and age determined the activity independence at hospital discharge. In addition, physiological and laboratory surveillance aided in predicting in-hospital deterioration. The use of the SHAP explanatory method successfully transformed machine learning predictions into clinically meaningful results.
- 25 March 2022
- Published by JMIR Publications Inc. in JMIR Medical Informatics
- Vol. 10 (3) , e32508
Open AccessMachine Learning Models for Predicting Influential Factors of Early Outcomes in Acute Ischemic Stroke: Registry-Based Study
Background: Timely and accurate outcome prediction plays a vital role in guiding clinical decisions on acute ischemic stroke. Early condition deterioration and severity after the acute stage are determinants for long-term outcomes. Therefore, predicting early outcomes is crucial in acute stroke management. However, interpreting the predictions and transforming them into clinically explainable concepts are as important as the predictions themselves. Objective: This work focused on machine learning model analysis in predicting the early outcomes of ischemic stroke and used model explanation skills in interpreting the results. Methods: Acute ischemic stroke patients registered on the Stroke Registry of the Chang Gung Healthcare System (SRICHS) in 2009 were enrolled for machine learning predictions of the two primary outcomes: modified Rankin Scale (mRS) at hospital discharge and in-hospital deterioration. We compared 4 machine learning models, namely support vector machine (SVM), random forest (RF), light gradient boosting machine (LGBM), and deep neural network (DNN), with the area under the curve (AUC) of the receiver operating characteristic curve. Further, 3 resampling methods, random under sampling (RUS), random over sampling, and the synthetic minority over-sampling technique, dealt with the imbalanced data. The models were explained based on the ranking of feature importance and the SHapley Additive exPlanations (SHAP). Results: RF performed well in both outcomes (discharge mRS: mean AUC 0.829, SD 0.018; in-hospital deterioration: mean AUC 0.710, SD 0.023 on original data and 0.728, SD 0.036 on resampled data with RUS for imbalanced data). In addition, DNN outperformed other models in predicting in-hospital deterioration on data without resampling (mean AUC 0.732, SD 0.064). In general, resampling contributed to the limited improvement of model performance in predicting in-hospital deterioration using imbalanced data. The features obtained from the National Institutes of Health Stroke Scale (NIHSS), white blood cell differential counts, and age were the key features for predicting discharge mRS. In contrast, the NIHSS total score, initial blood pressure, having diabetes mellitus, and features from hemograms were the most important features in predicting in-hospital deterioration. The SHAP summary described the impacts of the feature values on each outcome prediction. Conclusions: Machine learning models are feasible in predicting early stroke outcomes. An enriched feature bank could improve model performance. Initial neurological levels and age determined the activity independence at hospital discharge. In addition, physiological and laboratory surveillance aided in predicting in-hospital deterioration. The use of the SHAP explanatory method successfully transformed machine learning predictions into clinically meaningful results.
- Open Access
- 31 July 2021
- Published by JMIR Publications Inc. in JMIR Preprints
Machine Learning Models for Predicting Influential Factors of Early Outcomes in Acute Ischemic Stroke: Registry-Based Study (Preprint)
BACKGROUND Timely and accurate outcome prediction plays a vital role in guiding clinical decisions on acute ischemic stroke. Early condition deterioration and severity after the acute stage are determinants for long-term outcomes. Therefore, predicting early outcomes is crucial in acute stroke management. However, interpreting the predictions and transforming them into clinically explainable concepts are as important as the predictions themselves. OBJECTIVE This work focused on machine learning model analysis in predicting the early outcomes of ischemic stroke and used model explanation skills in interpreting the results. METHODS Acute ischemic stroke patients registered on the Stroke Registry of the Chang Gung Healthcare System (SRICHS) in 2009 were enrolled for machine learning predictions of the two primary outcomes: modified Rankin Scale (mRS) at hospital discharge and in-hospital deterioration. We compared 4 machine learning models, namely support vector machine (SVM), random forest (RF), light gradient boosting machine (LGBM), and deep neural network (DNN), with the area under the curve (AUC) of the receiver operating characteristic curve. Further, 3 resampling methods, random under sampling (RUS), random over sampling, and the synthetic minority over-sampling technique, dealt with the imbalanced data. The models were explained based on the ranking of feature importance and the SHapley Additive exPlanations (SHAP). RESULTS RF performed well in both outcomes (discharge mRS: mean AUC 0.829, SD 0.018; in-hospital deterioration: mean AUC 0.710, SD 0.023 on original data and 0.728, SD 0.036 on resampled data with RUS for imbalanced data). In addition, DNN outperformed other models in predicting in-hospital deterioration on data without resampling (mean AUC 0.732, SD 0.064). In general, resampling contributed to the limited improvement of model performance in predicting in-hospital deterioration using imbalanced data. The features obtained from the National Institutes of Health Stroke Scale (NIHSS), white blood cell differential counts, and age were the key features for predicting discharge mRS. In contrast, the NIHSS total score, initial blood pressure, having diabetes mellitus, and features from hemograms were the most important features in predicting in-hospital deterioration. The SHAP summary described the impacts of the feature values on each outcome prediction. CONCLUSIONS Machine learning models are feasible in predicting early stroke outcomes. An enriched feature bank could improve model performance. Initial neurological levels and age determined the activity independence at hospital discharge. In addition, physiological and laboratory surveillance aided in predicting in-hospital deterioration. The use of the SHAP explanatory method successfully transformed machine learning predictions into clinically meaningful results. CLINICALTRIAL
- 31 July 2021
- Published by JMIR Publications Inc. in JMIR Preprints
Open AccessMachine Learning Models for Predicting Influential Factors of Early Outcomes in Acute Ischemic Stroke: Registry-Based Study (Preprint)
BACKGROUND Timely and accurate outcome prediction plays a vital role in guiding clinical decisions on acute ischemic stroke. Early condition deterioration and severity after the acute stage are determinants for long-term outcomes. Therefore, predicting early outcomes is crucial in acute stroke management. However, interpreting the predictions and transforming them into clinically explainable concepts are as important as the predictions themselves. OBJECTIVE This work focused on machine learning model analysis in predicting the early outcomes of ischemic stroke and used model explanation skills in interpreting the results. METHODS Acute ischemic stroke patients registered on the Stroke Registry of the Chang Gung Healthcare System (SRICHS) in 2009 were enrolled for machine learning predictions of the two primary outcomes: modified Rankin Scale (mRS) at hospital discharge and in-hospital deterioration. We compared 4 machine learning models, namely support vector machine (SVM), random forest (RF), light gradient boosting machine (LGBM), and deep neural network (DNN), with the area under the curve (AUC) of the receiver operating characteristic curve. Further, 3 resampling methods, random under sampling (RUS), random over sampling, and the synthetic minority over-sampling technique, dealt with the imbalanced data. The models were explained based on the ranking of feature importance and the SHapley Additive exPlanations (SHAP). RESULTS RF performed well in both outcomes (discharge mRS: mean AUC 0.829, SD 0.018; in-hospital deterioration: mean AUC 0.710, SD 0.023 on original data and 0.728, SD 0.036 on resampled data with RUS for imbalanced data). In addition, DNN outperformed other models in predicting in-hospital deterioration on data without resampling (mean AUC 0.732, SD 0.064). In general, resampling contributed to the limited improvement of model performance in predicting in-hospital deterioration using imbalanced data. The features obtained from the National Institutes of Health Stroke Scale (NIHSS), white blood cell differential counts, and age were the key features for predicting discharge mRS. In contrast, the NIHSS total score, initial blood pressure, having diabetes mellitus, and features from hemograms were the most important features in predicting in-hospital deterioration. The SHAP summary described the impacts of the feature values on each outcome prediction. CONCLUSIONS Machine learning models are feasible in predicting early stroke outcomes. An enriched feature bank could improve model performance. Initial neurological levels and age determined the activity independence at hospital discharge. In addition, physiological and laboratory surveillance aided in predicting in-hospital deterioration. The use of the SHAP explanatory method successfully transformed machine learning predictions into clinically meaningful results. CLINICALTRIAL
- Open Access
- 6 April 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Emerging Topics in Computational Intelligence
- Vol. 4 (4) , 468-479
Unsupervised Representation Disentanglement Using Cross Domain Features and Adversarial Learning in Variational Autoencoder Based Voice Conversion
An effective approach for voice conversion (VC) is to disentangle linguistic content from other components in the speech signal. The effectiveness of variational autoencoder (VAE) based VC (VAE-VC), for instance, strongly relies on this principle. In our prior work, we proposed a cross-domain VAE-VC (CDVAE-VC) framework, which utilized acoustic features of different properties, to improve the performance of VAE-VC. We believed that the success came from more disentangled latent representations. In this article, we extend the CDVAE-VC framework by incorporating the concept of adversarial learning, in order to further increase the degree of disentanglement, thereby improving the quality and similarity of converted speech. More specifically, we first investigate the effectiveness of incorporating the generative adversarial networks (GANs) with CDVAE-VC. Then, we consider the concept of domain adversarial training and add an explicit constraint to the latent representation, realized by a speaker classifier, to explicitly eliminate the speaker information that resides in the latent code. Experimental results confirm that the degree of disentanglement of the learned latent representation can be enhanced by both GANs and the speaker classifier. Meanwhile, subjective evaluation results in terms of quality and similarity scores demonstrate the effectiveness of our proposed methods.
- 6 April 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Emerging Topics in Computational Intelligence
- Vol. 4 (4) , 468-479
Open AccessUnsupervised Representation Disentanglement Using Cross Domain Features and Adversarial Learning in Variational Autoencoder Based Voice Conversion
An effective approach for voice conversion (VC) is to disentangle linguistic content from other components in the speech signal. The effectiveness of variational autoencoder (VAE) based VC (VAE-VC), for instance, strongly relies on this principle. In our prior work, we proposed a cross-domain VAE-VC (CDVAE-VC) framework, which utilized acoustic features of different properties, to improve the performance of VAE-VC. We believed that the success came from more disentangled latent representations. In this article, we extend the CDVAE-VC framework by incorporating the concept of adversarial learning, in order to further increase the degree of disentanglement, thereby improving the quality and similarity of converted speech. More specifically, we first investigate the effectiveness of incorporating the generative adversarial networks (GANs) with CDVAE-VC. Then, we consider the concept of domain adversarial training and add an explicit constraint to the latent representation, realized by a speaker classifier, to explicitly eliminate the speaker information that resides in the latent code. Experimental results confirm that the degree of disentanglement of the learned latent representation can be enhanced by both GANs and the speaker classifier. Meanwhile, subjective evaluation results in terms of quality and similarity scores demonstrate the effectiveness of our proposed methods.