{"paper":{"title":"Approximate Distributed Coded Computing: Polynomial Codes and Randomized Sketching","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Combining polynomial codes and randomized sketching speeds up distributed optimization and machine learning despite slow servers.","cross_cats":["cs.IR","eess.SP"],"primary_cat":"cs.DC","authors_text":"Arya Mazumdar, Neophytos Charalambides","submitted_at":"2026-05-16T01:50:27Z","abstract_excerpt":"Coded computing is a distributed paradigm that uses coding theory to introduce \\textit{redundancy} and overcome bottlenecks in large-scale systems. In the same vein, randomized numerical linear algebra employs probabilistic methods to \\textit{compress} and accelerate linear algebraic operations, addressing challenges in high-dimensional data analysis. This article reviews the foundations of both fields and presents distributed schemes that combine techniques from both to speed up optimization and machine learning algorithms, in the presence of slow or non-responsive servers. Along the way, we "},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Distributed schemes that combine polynomial codes and randomized sketching can speed up optimization and machine learning algorithms in the presence of slow or non-responsive servers.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The integration of coding-based redundancy with sketching-based approximation preserves enough accuracy for the target optimization and learning tasks while still delivering net speedup.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Combines polynomial codes and randomized sketching into approximate distributed schemes that mitigate stragglers during optimization and machine learning tasks.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Combining polynomial codes and randomized sketching speeds up distributed optimization and machine learning despite slow servers.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"17d26b74a281d43e2bfc473e34c63163187f228a4f478a873d53934f3e84cb10"},"source":{"id":"2605.16744","kind":"arxiv","version":1},"verdict":{"id":"933a605b-7d22-4809-93af-697fdcb8b40e","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-19T19:55:39.360426Z","strongest_claim":"Distributed schemes that combine polynomial codes and randomized sketching can speed up optimization and machine learning algorithms in the presence of slow or non-responsive servers.","one_line_summary":"Combines polynomial codes and randomized sketching into approximate distributed schemes that mitigate stragglers during optimization and machine learning tasks.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The integration of coding-based redundancy with sketching-based approximation preserves enough accuracy for the target optimization and learning tasks while still delivering net speedup.","pith_extraction_headline":"Combining polynomial codes and randomized sketching speeds up distributed optimization and machine learning despite slow servers."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.16744/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"doi_title_agreement","ran_at":"2026-05-19T20:01:19.103044Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T20:01:13.642210Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"claim_evidence","ran_at":"2026-05-19T19:01:56.331797Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-19T18:33:26.461675Z","status":"skipped","version":"1.0.0","findings_count":0}],"snapshot_sha256":"d9a144ad31379d68de03c0a05ef8c74801ae35a4c6981353c3253f0df690b9e9"},"references":{"count":28,"sample":[{"doi":"","year":2018,"title":"Speeding Up Distributed Machine Learning Using Codes,","work_id":"1ac779d6-73d3-4a29-b67c-0db974d40eb2","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2020,"title":"Coded Computing: Miti- gating Fundamental Bottlenecks in Large-Scale Dis- tributed Computing and Machine Learning,","work_id":"6942de6c-d124-4b0c-9a05-f30cdb9ce6e4","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2015,"title":"A Practical Guide to Randomized Matrix Computations with MATLAB Implementations","work_id":"17c36ddd-0e99-4390-8f51-2da3fb5c175f","ref_index":3,"cited_arxiv_id":"1505.07570","is_internal_anchor":true},{"doi":"","year":2023,"title":"Ran- domized Numerical Linear Algebra: A Perspective on the Field With an Eye to Software,","work_id":"08fa85cd-7d23-44d6-a933-9dce4e5c5880","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2021,"title":"Numerically stable coded matrix computations via circulant and rota- tion matrix embeddings,","work_id":"3088a781-7e35-4e12-bce1-99d85aeed06e","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":28,"snapshot_sha256":"1da0b4418128d4d3e7435694e9ec6ac4918b2a08babcf250e62164045010123e","internal_anchors":2},"formal_canon":{"evidence_count":2,"snapshot_sha256":"a0b8fbb7b3406bb26313a545960a29e7a1b40336b65424af74d53fc573ab381b"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}