Brandeis University joins as 31st EdgeCloud customer, adopting EdgeCloud’s decentralized GPU…

Brandeis University joins as 31st EdgeCloud customer, adopting EdgeCloud’s decentralized GPU infrastructure for AI research

Theta is pleased to announce that the Liu Lab at Brandeis University, led by Professor Hongfu Liu, is adopting Theta EdgeCloud to accelerate its ML and AI research in data-centric learning, clustering analysis, and transfer learning. This partnership further establishes Theta’s leadership in the academia AI sector, making it the leading decentralized GPU infrastructure for researchers worldwide.

Professor Liu’s lab joins a growing number of leading academic institutions using EdgeCloud’s hybrid GPU infrastructure to enhance AI research productivity, including Stanford University, Seoul National University, KAIST, the University of Oregon, Michigan State University, NTU Singapore, and more.

By integrating EdgeCloud’s decentralized GPU resources into its research workflow, the Liu Lab gains access to scalable, high-performance, and cost-effective computing power, enabling faster development across a wide range of AI applications, enhancing the lab’s ability to innovate and push the boundaries of machine learning research.

“The integration of Theta EdgeCloud’s decentralized GPU infrastructure allows us to scale our experiments with ease, enabling our team to focus on research projects in machine learning and AI. The flexibility and cost-effectiveness of EdgeCloud’s infrastructure make it an invaluable asset for our research.” — Professor Hongfu Liu, Assistant Professor at Brandeis University

About Professor Hongfu Liu and Liu Lab

Professor Hongfu Liu is an Assistant Professor of Computer Science at Brandeis University. His research interests lie in core machine learning, with particular focuses on data-centric learning. He has authored numerous papers in top-tier conferences and journals and has received prestigious accolades, including the 2021 INNS Aharon Katzir Young Investigator Award, the 2022 Global Top Young Chinese Scholars in AI (Top 25 in Data Mining Area), and recognition as a highlighted/notable Area Chair at ICLR/NeurIPS 2022/2023. Dr. Liu also serves as an Associate Editor for IEEE CIM and TMLR and contributes as a (Senior) Area Chair for leading conferences such as ICLR, ICML, and NeurIPS for several years.

Data-Centric Learning: Recent Research at Liu Lab

Data-centric learning focuses on improving machine learning models by prioritizing the quality and diversity of the data used for training and evaluation, rather than exclusively refining algorithms. This approach recognizes that well-annotated, diverse, and representative datasets are crucial for building reliable and fair models. Data-centric learning is very similar to fitness, where it focuses on the diet, rather than exercise (See the picture below).

Liu Lab has conducted a series of studies on data-centric learning, with diverse applications in detrimental sample identification, noisy label correction, data trimming, data reweighting, active learning, poison attack and defense, alleviating distribution shift, LLM pre-training and fine-tuning, and so on.

Supporting AI Research with Theta EdgeCloud

Theta EdgeCloud’s decentralized GPU infrastructure empowers academic researchers with scalable, on-demand computing resources that are both high-performance and cost-effective. By offering access to a global network of GPUs, Theta EdgeCloud enables research teams, like the Liu Lab at Brandeis University, to efficiently accelerate their AI and machine learning projects. The hybrid nature of the infrastructure ensures that the Liu Lab can dynamically allocate computing resources, optimizing both performance and cost-efficiency for large-scale research initiatives.

By joining forces with institutions like Brandeis University, Theta Network will continue to support research projects with a cost-effective, easy onboarding process, enabling researchers to focus on various AI and machine learning research projects.

Brandeis University joins as 31st EdgeCloud customer, adopting EdgeCloud’s decentralized GPU… was originally published in Theta Network on Medium, where people are continuing the conversation by highlighting and responding to this story.