Chun-Hsi Huang

Chun-Hsi Huang

Auto-generated profile, based on publicly available information

Chun-Hsi Huang

School of Computing, Southern Illinois University, Carbondale, IL, USA

Research Focus & Metrics

Academic Career Length

23Years (2004-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
11 June 2025 in IEEE Transactions on Sustainable Computing
Verified
Journal Article
Closed access
SHSyed Mhamudul Hasan
Syed Mhamudul Hasan
Taminul IslamTaminul Islam
Taminul Islam
MSMunshi Saifuzzaman
Munshi Saifuzzaman
KAKhaled Sayed Ahmed
Khaled Sayed Ahmed
CHChun-Hsi Huang
Chun-Hsi Huang
Abdur Rahman Bin ShahidAbdur Rahman Bin Shahid
Abdur Rahman Bin Shahid
The rapid growth of machine learning (ML) technologies has raised significant concerns about their environmental impact, particularly regarding energy consumption and carbon emissions. This comprehensive review examines the intersection of ML and sustainability, synthesizing research from 2014 to 2024 to provide a holistic view of sustainable ML practices. This systematic review, encompassing over 200 peer-reviewed publications, reveals a growing emphasis on quantifying and mitigating the environmental footprint of ML systems. Key findings include: (1) a 300% increase in sustainable ML research since 2020; (2) the emergence of specialized carbon footprint quantification tools for ML; and (3) promising advancements in energy-efficient algorithms and green computing infrastructure. This research identifies critical challenges, including the lack of standardized sustainability metrics and the need for more robust life-cycle assessments of ML systems. The review also highlights the potential of transfer learning, federated learning, and hardware innovations in reducing ML's environmental impact. The analysis culminates in a novel framework for implementing sustainable practices in ML projects and a detailed roadmap for future research. This work provides researchers, practitioners, and policymakers with crucial insights to drive the development of more environmentally responsible ML technologies, ultimately contributing to global sustainability goals.

1 December 2016
Verified
Conference Paper
Closed access
HGHaitham Ghalwash
Haitham Ghalwash
CHChun-Hsi Huang
Chun-Hsi Huang
Significant research effort has been made on formulating new topologies to meet the requirements of current and future large-scale data centers. Nowadays data centers may include tens of thousands of servers, leading to an urgent need for higher bandwidth, better reliability, easier management and lower latency. This paper investigates the potential of using software-defined networks (SDNs) in Extreme scale systems, which are required for future data centers when scaling up to the Exascale level. SDN is the new trend for efficient traffic control and network management. We investigate the SDN-based behavior and performance of two widely used topologies in large-scale data centers, namely Fat-Tree and BCube, using different network scales, traffic loads and traffic patterns. Experimental results based on such performance metrics as delay, jitter and packet drops have indicated that the BCube-like topology outperforms the Fat-Tree topology in small-scale networks, while the Fat-Tree performs better as the network scales up.

2 February 2019
Verified
Book Chapter
Open access
PDF file
HGHaitham Ghalwash
Haitham Ghalwash
CHChun-Hsi Huang
Chun-Hsi Huang
Software-defined Networks (SDNs) are the new network paradigm providing, programmability, agility, and centralized management. In this paper, we show how to leverage the SDN centralized controller to improve the network utilization and the traffic performance. On top of the SDN controller, new modules are added to help finding single and multi-path routes between communicating devices. Flow rules are automatically installed into the designated switches to provide the required paths. The behavior and performance of different types of traffic, namely, UDP, TCP, VOIP, and a Big-data application traffic are investigated. The traffic forwarding is based on either the controller built in layer 2 switching “odl-l2switch” feature or single/multi-path selection based on the supplemented modules. Experimental results based on metrics such as delay, jitter and packet drops are presented for each forwarding option. The results disclosed the advantage of having the developed modules on top of the controller for all traffic types. The OpenDaylight controller for OpenFlow switches, in a fat-tree network, is used for experiments. For a fair comparison of different traffic types, a monitoring module is built on top of the controller for collecting ports statistics, analyzing and monitoring.

2 March 2004
Verified
Conference Paper
Closed access
CLChain-Wu Lee
Chain-Wu Lee
CHChun-Hsi Huang
Chun-Hsi Huang
SRS. Rajasekaran
S. Rajasekaran

1 March 2019
Verified
Conference Paper
Closed access
BABadar Almarri
Badar Almarri
CHChun-Hsi Huang
Chun-Hsi Huang
Ground truth labels are expected to be certain, and their existence is essentially a vital component of supervised learning problems. In certain cases, however, they can prove to be obstacles. They can lead to two possible issues: class imbalances due to skewed label distributions, and unreliability due to the uncertainty of raters underlying rationale. In cases where the labels are continuous, they need to be dichotomized for a classification task. Dichotomization is often decided statistically or based on the subject matter. However, the subjectivity of participants and its impact is neglected. In this paper, we investigate the effect of thresholding on an EEG emotional self-assessment. We propose a modification in the prediction pipeline to minimize subjectivity, improving model outcomes as well as alleviating the effect of label imbalance.