{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,6]],"date-time":"2026-05-06T15:20:30Z","timestamp":1778080830316,"version":"3.51.4"},"reference-count":50,"publisher":"Association for Computing Machinery (ACM)","issue":"4","license":[{"start":{"date-parts":[[2016,5,2]],"date-time":"2016-05-02T00:00:00Z","timestamp":1462147200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/www.acm.org\/publications\/policies\/copyright_policy#Background"}],"funder":[{"name":"\u201cMomentum\u2014Big Data\u201d grant of the Hungarian Academy of Sciences"},{"name":"OTKA NK","award":["105645"],"award-info":[{"award-number":["105645"]}]}],"content-domain":{"domain":["dl.acm.org"],"crossmark-restriction":true},"short-container-title":["ACM Trans. Intell. Syst. Technol."],"published-print":{"date-parts":[[2016,7,14]]},"abstract":"<jats:p>Low-rank matrix approximation is an important tool in data mining with a wide range of applications, including recommender systems, clustering, and identifying topics in documents. When the matrix to be approximated originates from a large distributed system, such as a network of mobile phones or smart meters, a challenging problem arises due to the strongly conflicting yet essential requirements of efficiency, robustness, and privacy preservation. We argue that although collecting sensitive data in a centralized fashion may be efficient, it is not an option when considering privacy and efficiency at the same time. Thus, we do not allow any sensitive data to leave the nodes of the network. The local information at each node (personal attributes, documents, media ratings, etc.) defines one row in the matrix. This means that all computations have to be performed at the edge of the network. Known parallel methods that respect the locality constraint, such as synchronized parallel gradient search or distributed iterative methods, require synchronized rounds or have inherent issues with load balancing, and thus they are not robust to failure. Our distributed stochastic gradient descent algorithm overcomes these limitations. During the execution, any sensitive information remains local, whereas the global features (e.g., the factor model of movies) converge to the correct value at all nodes. We present a theoretical derivation and a thorough experimental evaluation of our algorithm. We demonstrate that the convergence speed of our method is competitive while not relying on synchronization and being robust to extreme and realistic failure scenarios. To demonstrate the feasibility of our approach, we present trace-based simulations, real smartphone user behavior analysis, and tests over real movie recommender system data.<\/jats:p>","DOI":"10.1145\/2854157","type":"journal-article","created":{"date-parts":[[2016,5,3]],"date-time":"2016-05-03T13:08:37Z","timestamp":1462280917000},"page":"1-24","update-policy":"https:\/\/doi.org\/10.1145\/crossmark-policy","source":"Crossref","is-referenced-by-count":17,"title":["Robust Decentralized Low-Rank Matrix Decomposition"],"prefix":"10.1145","volume":"7","author":[{"given":"Istv\u00e1n","family":"Heged\u0171s","sequence":"first","affiliation":[{"name":"University of Szeged, Hungary"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"\u00c1rp\u00e1d","family":"Berta","sequence":"additional","affiliation":[{"name":"University of Szeged, Hungary"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Levente","family":"Kocsis","sequence":"additional","affiliation":[{"name":"Institute for Computer Science and Control, Hungarian Academy of Sciences (MTA SZTAKI)"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Andr\u00e1s A.","family":"Bencz\u00far","sequence":"additional","affiliation":[{"name":"Institute for Computer Science and Control, Hungarian Academy of Sciences (MTA SZTAKI)"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"M\u00e1rk","family":"Jelasity","sequence":"additional","affiliation":[{"name":"University of Szeged, and MTA-SZTE Research Group on AI, Hungary"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"320","published-online":{"date-parts":[[2016,5,2]]},"reference":[{"key":"e_1_2_1_1_1","doi-asserted-by":"publisher","DOI":"10.1007\/11503415_31"},{"key":"e_1_2_1_2_1","volume-title":"Proceedings of the IEEE Global Telecommunications Conference (GLOBECOM\u201906)","author":"Ahmad Waseem","year":"2006"},{"key":"e_1_2_1_3_1","volume-title":"Introduction to Machine Learning","author":"Alpaydin Ethem","edition":"2"},{"key":"e_1_2_1_4_1","doi-asserted-by":"publisher","DOI":"10.1145\/380752.380859"},{"key":"e_1_2_1_5_1","volume-title":"UCI Machine Learning Repository. Retrieved","author":"Bache K.","year":"2016"},{"key":"e_1_2_1_6_1","doi-asserted-by":"crossref","unstructured":"Austin R. Benson David F. Gleich and James Demmel. 2013. Direct QR factorizations for tall-and-skinny matrices in MapReduce architectures. arXiv:1301.1071 {cs.DC}.  Austin R. Benson David F. Gleich and James Demmel. 2013. Direct QR factorizations for tall-and-skinny matrices in MapReduce architectures. arXiv:1301.1071 {cs.DC}.","DOI":"10.1109\/BigData.2013.6691583"},{"key":"e_1_2_1_7_1","doi-asserted-by":"publisher","DOI":"10.1145\/2792838.2800173"},{"key":"e_1_2_1_8_1","doi-asserted-by":"publisher","DOI":"10.1137\/1037127"},{"key":"e_1_2_1_9_1","doi-asserted-by":"publisher","DOI":"10.1109\/P2P.2014.6934317"},{"key":"e_1_2_1_10_1","doi-asserted-by":"publisher","DOI":"10.1145\/2723872.2723891"},{"key":"e_1_2_1_11_1","volume-title":"Yi-An Lin, YuanYuan Yu, Gary Bradski, Andrew Y. Ng, and Kunle Olukotun.","author":"Chu Cheng-Tao","year":"2007"},{"key":"e_1_2_1_12_1","doi-asserted-by":"publisher","DOI":"10.1007\/s000260300002"},{"key":"e_1_2_1_13_1","series-title":"Lecture Notes in Computer Science","volume-title":"Distributed Applications and Interoperable Systems","author":"Danner G\u00e1bor"},{"key":"e_1_2_1_14_1","doi-asserted-by":"publisher","DOI":"10.1023\/B:MACH.0000033113.59016.96"},{"key":"e_1_2_1_15_1","doi-asserted-by":"publisher","DOI":"10.1137\/S0097539704442696"},{"key":"e_1_2_1_16_1","doi-asserted-by":"publisher","DOI":"10.1145\/509907.509922"},{"key":"e_1_2_1_17_1","doi-asserted-by":"publisher","DOI":"10.1145\/1866739.1866758"},{"key":"e_1_2_1_18_1","doi-asserted-by":"publisher","DOI":"10.1145\/2020408.2020426"},{"key":"e_1_2_1_19_1","doi-asserted-by":"crossref","volume-title":"Methods for generating random orthogonal matrices","author":"Genz Alan","DOI":"10.1007\/978-3-642-59657-5_13"},{"key":"e_1_2_1_20_1","volume-title":"Proceedings of the 11th Conference of the European Chapter of the Association for Computational Linguistics (EACL\u201906)","author":"Gorrell Genevieve","year":"2006"},{"key":"e_1_2_1_21_1","doi-asserted-by":"publisher","DOI":"10.1109\/TSP.2012.2190406"},{"key":"e_1_2_1_22_1","doi-asserted-by":"publisher","DOI":"10.1109\/TNNLS.2012.2197827"},{"key":"e_1_2_1_23_1","doi-asserted-by":"publisher","DOI":"10.1145\/2684464.2684480"},{"key":"e_1_2_1_24_1","doi-asserted-by":"publisher","DOI":"10.1145\/2043932.2043948"},{"key":"e_1_2_1_25_1","doi-asserted-by":"publisher","DOI":"10.1007\/11503415_30"},{"key":"e_1_2_1_26_1","doi-asserted-by":"publisher","DOI":"10.1145\/1007352.1007438"},{"key":"e_1_2_1_27_1","doi-asserted-by":"publisher","DOI":"10.1145\/324133.324140"},{"key":"e_1_2_1_28_1","doi-asserted-by":"publisher","DOI":"10.1145\/1993744.1993764"},{"key":"e_1_2_1_29_1","doi-asserted-by":"publisher","DOI":"10.1109\/MC.2009.263"},{"key":"e_1_2_1_30_1","volume-title":"Proceedings of the 29th International Conference on Machine Learning (ICML\u201912)","author":"Le Quoc","year":"2012"},{"key":"e_1_2_1_31_1","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-642-12963-6_2"},{"key":"e_1_2_1_32_1","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2012.6288528"},{"key":"e_1_2_1_33_1","volume-title":"Proceedings of the Conference on Uncertainty in Artificial Intelligence.","author":"Low Yucheng"},{"key":"e_1_2_1_34_1","doi-asserted-by":"publisher","DOI":"10.5555\/874063.875554"},{"key":"e_1_2_1_35_1","series-title":"Lecture Notes in Computer Science","volume-title":"Randomization and Approximation Techniques in Computer Science","author":"Mihail Milena"},{"key":"e_1_2_1_36_1","doi-asserted-by":"publisher","DOI":"10.1109\/P2P.2009.5284506"},{"key":"e_1_2_1_37_1","doi-asserted-by":"publisher","DOI":"10.1145\/2508859.2516751"},{"key":"e_1_2_1_38_1","volume-title":"Retrieved","author":"Nis T.","year":"1999"},{"key":"e_1_2_1_39_1","doi-asserted-by":"publisher","DOI":"10.1002\/cpe.2858"},{"key":"e_1_2_1_40_1","doi-asserted-by":"publisher","DOI":"10.1006\/jcss.2000.1711"},{"key":"e_1_2_1_41_1","doi-asserted-by":"publisher","DOI":"10.1145\/2645710.2645725"},{"key":"e_1_2_1_42_1","doi-asserted-by":"publisher","DOI":"10.1145\/192844.192905"},{"key":"e_1_2_1_43_1","volume-title":"Secure Decentralized Swarm Discovery in Tribler. Master\u2019s Thesis","author":"Roozenburg Jelle"},{"key":"e_1_2_1_44_1","doi-asserted-by":"publisher","DOI":"10.1109\/P2P.2013.6688707"},{"key":"e_1_2_1_45_1","volume-title":"Proceedings of the 20th International Conference on Machine Learning (ICML\u201903)","author":"Srebro Nathan","year":"2003"},{"key":"e_1_2_1_46_1","doi-asserted-by":"publisher","DOI":"10.1145\/1177080.1177105"},{"key":"e_1_2_1_47_1","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-642-03869-3_50"},{"key":"e_1_2_1_48_1","volume-title":"Proceedings of the 32nd International Conference on Machine Learning (ICML\u201915)","author":"Wang Yu-Xiang","year":"2015"},{"key":"e_1_2_1_49_1","doi-asserted-by":"publisher","DOI":"10.1109\/TKDE.2012.191"},{"key":"e_1_2_1_50_1","unstructured":"Martin A. Zinkevich Alex Smola Markus Weimer and Lihong Li. 2010. Parallelized stochastic gradient descent. In Advances in Neural Information Processing Systems 23 (NIPS\u201910). 2595--2603.  Martin A. Zinkevich Alex Smola Markus Weimer and Lihong Li. 2010. Parallelized stochastic gradient descent. In Advances in Neural Information Processing Systems 23 (NIPS\u201910). 2595--2603."}],"container-title":["ACM Transactions on Intelligent Systems and Technology"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/2854157","content-type":"unspecified","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.1145\/2854157","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,6,18]],"date-time":"2025-06-18T05:48:47Z","timestamp":1750225727000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/2854157"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2016,5,2]]},"references-count":50,"journal-issue":{"issue":"4","published-print":{"date-parts":[[2016,7,14]]}},"alternative-id":["10.1145\/2854157"],"URL":"https:\/\/doi.org\/10.1145\/2854157","relation":{},"ISSN":["2157-6904","2157-6912"],"issn-type":[{"value":"2157-6904","type":"print"},{"value":"2157-6912","type":"electronic"}],"subject":[],"published":{"date-parts":[[2016,5,2]]},"assertion":[{"value":"2015-03-01","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2015-12-01","order":1,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2016-05-02","order":2,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}