Skip to content

Commit

Permalink
Update publications and projects
Browse files Browse the repository at this point in the history
  • Loading branch information
xuchengustb committed Dec 7, 2024
1 parent 3c6f31f commit 63a663d
Show file tree
Hide file tree
Showing 14 changed files with 463 additions and 9 deletions.
Binary file modified .DS_Store
Binary file not shown.
6 changes: 5 additions & 1 deletion docs/project/index.html
Original file line number Diff line number Diff line change
Expand Up @@ -113,7 +113,11 @@ <h1 class="post__title">Projects</h1>
<p>This page is still under construction !</p>
<!-- raw HTML omitted -->
<h1 id="matswarmhahahugoshortcode7s1hbhb"><a href="/projects/matswarm/">MatSwarm</a></h1>
<p>The rapid evolution of Industry 4.0 demands seamless collaboration among material research institutions to speed up advanced material discovery. The current platforms struggle with integrating large-scale, heterogeneous datasets, leading to data silos that hinder collaboration and innovation. The University of Science and Technology Beijing addresses these challenges through the National Material Data Management and Services (NMDMS) platform, which aggregates over 14 million material data entries from 30+ institutions, supporting high-throughput experimentation and collaborative research in materials genomic engineering. Key to this platform&rsquo;s success is its advanced data normalization, distributed storage, and blockchain-based middleware, which ensures secure, cross-institutional data sharing. The MatSwarm framework further enhances this environment by introducing swarm transfer learning to boost model accuracy and generalization on non-i.i.d. data. NMDMS stands as a pioneering tool in materials research, driving innovation and fostering secure, efficient, and collaborative materials computation across institutions.</p>
<p><img src="/img/MatSwarm.png" alt="system">
The rapid evolution of Industry 4.0 demands seamless collaboration among material research institutions to speed up advanced material discovery. The current platforms struggle with integrating large-scale, heterogeneous datasets, leading to data silos that hinder collaboration and innovation. The University of Science and Technology Beijing addresses these challenges through the National Material Data Management and Services (NMDMS) platform, which aggregates over 14 million material data entries from 30+ institutions, supporting high-throughput experimentation and collaborative research in materials genomic engineering. Key to this platform&rsquo;s success is its advanced data normalization, distributed storage, and blockchain-based middleware, which ensures secure, cross-institutional data sharing. The MatSwarm framework further enhances this environment by introducing swarm transfer learning to boost model accuracy and generalization on non-i.i.d. data. NMDMS stands as a pioneering tool in materials research, driving innovation and fostering secure, efficient, and collaborative materials computation across institutions.</p>
<h1 id="fedmdhhahahugoshortcode7s2hbhb"><a href="/projects/fedmdh/">FedMDH</a></h1>
<p><img src="/img/FedMDH.png" alt="system">
In the field of materials science, due to various factors such as material sources, testing equipment, and technical methods, the data distributions across different organizations are often non-identical and non-independent (non-i.i.d.) . This data heterogeneity can manifest in various forms, including 1) feature space disparity, 2) sample imbalance, and 3) label distribution variance. We define it as multi-dimensional heterogeneity (MDH). To overcome these challenges, we introduce FedMDH, a federated learning framework designed to tackle Multi-Dimensional Heterogeneity. While FedMDH is applicable to various downstream tasks, this work focuses on the widespread, complex, and underexplored regression tasks in materials science.</p>
<!-- raw HTML omitted -->
<!-- raw HTML omitted -->

Expand Down
9 changes: 5 additions & 4 deletions docs/publications/index.html
Original file line number Diff line number Diff line change
Expand Up @@ -115,16 +115,17 @@ <h1 class="post__title">Publications</h1>
<h2 id="acceptedonlinearxiv">Accepted/Online/Arxiv</h2>
<ol>
<li>
<p>Wang R, Ma F*, Tang S, Zhang H, He J, Su Z, Zhang X, <strong>Xu C*</strong>. Parallel Byzantine Fault Tolerance Consensus
Based on Trusted Execution Environments[J]. <strong><em>Peer-to-Peer Networking and Applications</em></strong>, 2024. <strong><em>Accepted</em></strong></p>
</li>
<li>
<p><strong>Xu C</strong>, Zhang C, Shi Y, Wang R, Duan S*, Wan Y*, Zhang X. Subgoal-based Hierarchical Reinforcement Learning for Multi-Agent Collaboration [EB/OL]. <strong><em>Under Review</em></strong>, arXiv:2408.11416. Available from: <a href="https://arxiv.org/abs/2408.11416">https://arxiv.org/abs/2408.11416</a></p>
</li>
<li>
<p>Shi Y, Duan S, <strong>Xu C*</strong>, Wang R, Ye F. &amp; Yuen, C. (2024). Dynamic Deep Factor Graph for Multi-Agent Reinforcement Learning. <strong><em>Under Review</em></strong>, arXiv:2405.05542. Available from <a href="https://arxiv.org/abs/2405.05542">https://arxiv.org/abs/2405.05542</a></p>
</li>
</ol>
<h2 id="2025">2025</h2>
<ol>
<li>Wang R, Ma F*, Tang S, Zhang H, He J, Su Z, Zhang X, <strong>Xu C*</strong>. Parallel Byzantine Fault Tolerance Consensus
Based on Trusted Execution Environments[J]. <strong><em>Peer-to-Peer Networking and Applications</em></strong>, 2025, 18:1-24. <a href="/pdf/2025_Wang_Parallel%20Byzantine%20fault%20tolerance%20consensus%20based%20on%20trusted%20execution%20environments.pdf">PDF</a></li>
</ol>
<h2 id="2024">2024</h2>
<ol>
<li>
Expand Down
2 changes: 1 addition & 1 deletion docs/publications_ds/index.html
Original file line number Diff line number Diff line change
Expand Up @@ -125,7 +125,7 @@ <h1 class="post__title">Publications on Topic of Distributed Scurity</h1>
</li>
<li>
<p>Wang R, Ma F*, Tang S, Zhang H, He J, Su Z, Zhang X, <strong>Xu C*</strong>. Parallel Byzantine Fault Tolerance Consensus
Based on Trusted Execution Environments[J]. <strong><em>Peer-to-Peer Networking and Applications</em></strong>, 2024. <strong><em>Accepted</em></strong></p>
Based on Trusted Execution Environments[J]. <strong><em>Peer-to-Peer Networking and Applications</em></strong>, 2025, 18:1-24. <a href="/pdf/2025_Wang_Parallel%20Byzantine%20fault%20tolerance%20consensus%20based%20on%20trusted%20execution%20environments.pdf">PDF</a></p>
</li>
</ol>

Expand Down
Binary file added img/FedMDH.png
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file added img/MatSwarm.png
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
7 changes: 7 additions & 0 deletions index.xml
Original file line number Diff line number Diff line change
Expand Up @@ -85,6 +85,13 @@ Wan J, Xu C*, Chen W, Wang R, Zhang X*. Abrupt moving target tracking based on q
Wang R, Xu C*, Zhang X*. Toward Materials Genome Big-Data: A Blockchain-based Secure Storage and Efficient Retrieval Method[J]. IEEE Transactions on Parallel and Distributed Systems, 2024, 35(9): 1630–1643. PDF
Wang R, Xu C*, Ye F, Tang S, Zhang X*.</description>
</item>
<item>
<title></title>
<link>/projects/fedmdh/</link>
<pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate>
<guid>/projects/fedmdh/</guid>
<description>FedMDH: A Federated Learning Framework for Effective Sharing of Multi-Dimensional Heterogeneous Materials Data In the field of materials science, due to various factors such as material sources, testing equipment, and technical methods, the data distributions across different organizations are often non-identical and non-independent (non-i.i.d.) . This data heterogeneity can manifest in various forms, including 1) feature space disparity, 2) sample imbalance, and 3) label distribution variance. We define it as multi-dimensional heterogeneity (MDH).</description>
</item>
<item>
<title></title>
<link>/projects/matswarm/</link>
Expand Down
Binary file not shown.
Loading

0 comments on commit 63a663d

Please sign in to comment.