{"691036":{"#nid":"691036","#data":{"type":"event","title":"MS Defense by Tiancheng Zhao","body":[{"value":"\u003Cp\u003E\u0026nbsp;\u003C\/p\u003E\u003Cp\u003EStudent Information:\u0026nbsp;\u003C\/p\u003E\u003Cp\u003E\u2002\u2002\u2002\u2002Full Name: Tiancheng Zhao\u003C\/p\u003E\u003Cp\u003E\u2002\u2002\u2002\u2002GT ID: 904056610\u003C\/p\u003E\u003Cp\u003E\u2002\u2002\u2002\u2002Degree: Master of Science in Computer Science\u003C\/p\u003E\u003Cp\u003E\u2002\u2002\u2002\u2002School: School of Computer Science, College of Computing\u003C\/p\u003E\u003Cp\u003E\u0026nbsp;\u003C\/p\u003E\u003Cp\u003E\u2002\u2002\u2002\u2002Defense Schedule \u0026amp; Location:\u003C\/p\u003E\u003Cp\u003E\u2002\u2002\u2002\u2002\u2002\u2002\u2002\u2002Date: Tuesday, July 14, 2026\u003C\/p\u003E\u003Cp\u003E\u2002\u2002\u2002\u2002\u2002\u2002\u2002\u2002Time: 10.30 am - 12.30 pm\u003C\/p\u003E\u003Cp\u003E\u2002\u2002\u2002\u2002\u2002\u2002\u2002\u2002Location: \u003Cstrong\u003ECoda\u003C\/strong\u003E\u0026nbsp;\u003Cstrong\u003EC1315 Grant Park\u003C\/strong\u003E; \u003Ca href=\u0022https:\/\/gatech.zoom.us\/j\/98984363801?pwd=2Nho0mHrZYgO4Of\u0022\u003E\u003Cstrong\u003Ehttps:\/\/gatech.zoom.us\/j\/98984363801?pwd=2Nho0mHrZYgO4Of\u003C\/strong\u003E\u003C\/a\u003E\u003C\/p\u003E\u003Cp\u003E\u0026nbsp;\u003C\/p\u003E\u003Cp\u003E\u2002\u2002\u2002Thesis Details:\u003C\/p\u003E\u003Cp\u003E\u2002\u2002\u2002\u2002\u2002\u2002\u2002\u2002Title: A Fast Parallel Density-Aware Algorithm for Graph Spectral Sparsification\u003C\/p\u003E\u003Cp\u003E\u0026nbsp;\u003C\/p\u003E\u003Cp\u003E\u2002\u2002\u2002\u2002\u003Cstrong\u003ECommittee Members:\u003C\/strong\u003E\u003C\/p\u003E\u003Cp\u003E\u2022 \u0026nbsp;\u0026nbsp; Prof. Helen Xu (Advisor) - School of Computational Science and Engineering, Georgia Institute of Technology\u003C\/p\u003E\u003Cp\u003E\u2022 \u0026nbsp;\u0026nbsp; Prof. Richard Vuduc - School of Computational Science and Engineering, Georgia Institute of Technology\u003C\/p\u003E\u003Cp\u003E\u2022 \u0026nbsp;\u0026nbsp; Prof.\u0026nbsp;Jan van den Brand \u2013 School of Computer Science, Georgia Institute of Technology\u003C\/p\u003E\u003Cp\u003E\u2022 \u0026nbsp; \u0026nbsp;Prof. Elizabeth Cherry\u0026nbsp;\u2013 School of Computer Science, Georgia Institute of Technology\u003C\/p\u003E\u003Cp\u003E\u2022 \u0026nbsp;\u0026nbsp; Prof.\u0026nbsp;Chaojian Li - Department of Computer Science and Engineering, Hong Kong University of Science and Technology\u003C\/p\u003E\u003Cp\u003E\u0026nbsp;\u003C\/p\u003E\u003Cp\u003E\u003Cstrong\u003E\u2002\u2002\u2002\u2002Abstract\u003C\/strong\u003E\u003C\/p\u003E\u003Cp\u003E\u0026nbsp;\u003C\/p\u003E\u003Cp\u003EGraph Spectral Sparsification (GSS) identifies an ultra-sparse subgraph, or sparsifier,\u0026nbsp;whose Laplacian matrix closely approximates the spectral properties of the original graph,\u0026nbsp;enabling substantial reductions in computational complexity for computationally intensive\u0026nbsp;problems in scientific computing.\u0026nbsp;The state-of-the-art method for efficient GSS is feGRASS,\u0026nbsp;consisting of two steps: 1) spanning tree generation and 2) off-tree edge recovery. However,\u0026nbsp;feGRASS suffers from two main issues: 1) difficulties in parallelizing the recovery step for\u0026nbsp;strict data dependencies, and 2) performance degradation on skewed inputs, often requiring\u0026nbsp;multiple passes to recover sufficient edges.\u0026nbsp;To address these challenges, we propose parallel density-aware Graph Spectral Sparsification (pdGRASS), a parallel algorithm that organizes edges into disjoint subtasks without\u0026nbsp;data dependencies between them, enabling efficient parallelization and sufficient edge recovery in a single pass. We empirically evaluate feGRASS and pdGRASS based on 1)\u0026nbsp;off-tree edge-recovery runtime and 2) sparsifier quality, measured by the iteration count\u0026nbsp;required for convergence in a preconditioned conjugate gradient (PCG) application. The\u0026nbsp;evaluation demonstrates that pdGRASS achieves average speedups ranging from 3.9\u00d7 to\u0026nbsp;8.8\u00d7 depending on the number of edges recovered, and mitigates worst-case runtimes of\u0026nbsp;feGRASS with over 1000\u00d7 speedup on skewed inputs.\u003C\/p\u003E","summary":"","format":"limited_html"}],"field_subtitle":"","field_summary":[{"value":"\u003Cp\u003EA Fast Parallel Density-Aware Algorithm for Graph Spectral Sparsification\u003C\/p\u003E","format":"limited_html"}],"field_summary_sentence":[{"value":"A Fast Parallel Density-Aware Algorithm for Graph Spectral Sparsification"}],"uid":"27707","created_gmt":"2026-07-06 13:31:45","changed_gmt":"2026-07-06 13:32:29","author":"Tatianna Richardson","boilerplate_text":"","field_publication":"","field_article_url":"","field_event_time":{"event_time_start":"2026-07-14T10:30:00-04:00","event_time_end":"2026-07-14T12:30:00-04:00","event_time_end_last":"2026-07-14T12:30:00-04:00","gmt_time_start":"2026-07-14 14:30:00","gmt_time_end":"2026-07-14 16:30:00","gmt_time_end_last":"2026-07-14 16:30:00","rrule":null,"timezone":"America\/New_York"},"location":"Coda C1315 Grant Park","extras":[],"groups":[{"id":"221981","name":"Graduate Studies"}],"categories":[],"keywords":[{"id":"111531","name":"ms defense"}],"core_research_areas":[],"news_room_topics":[],"event_categories":[{"id":"1788","name":"Other\/Miscellaneous"}],"invited_audience":[{"id":"78771","name":"Public"}],"affiliations":[],"classification":[],"areas_of_expertise":[],"news_and_recent_appearances":[],"phone":[],"contact":[],"email":[],"slides":[],"orientation":[],"userdata":""}}}