6 Hits in 2.3 sec

FlexiFaCT: Scalable Flexible Factorization of Coupled Tensors on Hadoop [chapter]

Alex Beutel, Partha Pratim Talukdar, Abhimanu Kumar, Christos Faloutsos, Evangelos E. Papalexakis, Eric P. Xing
2014 Proceedings of the 2014 SIAM International Conference on Data Mining  
In Proceedings of the 17th ACM SIGKDD international conference on Knowledge discovery and data mining, KDD 11 [2] Martin Zinkevich, Markus Weimer, Lihong Li, and Alex J Smola.  ...  Gigatensor: scaling tensor analysis up by 100 timesalgorithms and discoveries. In KDD' 12 Factoring your data into its component pieces can be insightful for data mining, and useful for prediction.  ...  Scalability In Figure 2 (b) we show how coupled factorization using FLEXIFACT scales, as the dimensions of the data increase.  ... 
doi:10.1137/1.9781611973440.13 dblp:conf/sdm/BeutelTKFPX14 fatcat:gjqmuletxvdsfcsqwwbeeip2jq

FlexiFaCT: Scalable Flexible Factorization of Coupled Tensors on Hadoop

Alex Beutel, Abhimanu Kumar, Evangelos E. Papalexakis, Partha Pratim Talukdar, Christos Faloutsos, Eric P Xing
FlexiFaCT runs on standard Hadoop. (3) Convergence proofs showing that Flexi-FaCT converges on the variety of objective functions, even with projections.  ...  We provide a distributed, scalable method for decomposing matrices, tensors, and coupled data sets through stochastic gradient descent on a variety of objective functions.  ...  Last, we would like to thank the Open Cloud Consortium (OCC) and the Open Science Data Cloud (OSDC) for the use of resources on the OCC-Y Hadoop cluster, which was donated to the OCC by Yahoo!  ... 
doi:10.1184/r1/6475637 fatcat:r6zivuqqj5ejrbsjappxacsqka

SCouT: Scalable coupled matrix-tensor factorization - algorithm and discoveries

ByungSoo Jeon, Inah Jeon, Lee Sael, U Kang
2016 2016 IEEE 32nd International Conference on Data Engineering (ICDE)  
In this paper, we propose SCOUT, a large-scale coupled matrix-tensor factorization algorithm running on the distributed MAPREDUCE platform.  ...  However, existing single machine or distributed algorithms for coupled matrix-tensor factorization do not scale for tensors with billions of elements in each mode.  ...  [7] propose FlexiFaCT, a flexible tensor decomposition method based on distributed stochastic gradient descent.  ... 
doi:10.1109/icde.2016.7498292 dblp:conf/icde/JeonJSK16 fatcat:vvvy7uqznrbdhgabizbln4s62m

J-Recs: Principled and Scalable Recommendation Justification [article]

Namyong Park, Andrey Kan, Christos Faloutsos, Xin Luna Dong
2020 arXiv   pre-print
Existing post-hoc methods are often limited in providing diverse justifications, as they either use only one of many available types of input data, or rely on the predefined templates.  ...  J-Recs is a recommendation model-agnostic method that generates diverse justifications based on various types of product and user data (e.g., purchase history and product attributes).  ...  recovery Next-generation phenotyping of electronic health records FlexiFaCT: scalable flexible factorization of coupled tensors on Hadoop FlexiFaCT: scalable flexible factorization of coupled tensors  ... 
arXiv:2011.05928v1 fatcat:3dqm6aqazzbc5fcmwiapv7ewza

Modeling Large Social Networks in Context

Qirong Ho
on statistical models, and (2) strategies for network data representation, model design, algorithm design and distributed multi-machine programming that, together, ensure scalability to very large networks  ...  The methods presented herein combine the flexibility of statistical models with key ideas and empirical observations from the data mining and social networks communities, and are supported by software  ...  coupled MF, such as FlexiFaCT [18] .  ... 
doi:10.1184/r1/6720683.v1 fatcat:rr6jq6ssbjefrkq4fpybc7sppa

Diversity-Promoting and Large-Scale Machine Learning for Healthcare

Pengtao Xie
We provide theoretical analysis on why promoting diversity canbetter capture [...]  ...  When developing these models, we encounter several challenges: (1) How to better capture infrequent clinical patterns,such as rare subtypes of diseases; (2) How to make the models generalize well on unseen  ...  The tree LSTMs are built on the constituent parse trees of individual sentences and the sequential LSTM is built on the sequence of sentences.  ... 
doi:10.1184/r1/7553468 fatcat:ac5ifp2lnzbk3hcupr2rszxj2m