67,750 Hits in 5.1 sec

Local collaborative ranking

Joonseok Lee, Samy Bengio, Seungyeon Kim, Guy Lebanon, Yoram Singer
2014 Proceedings of the 23rd international conference on World wide web - WWW '14  
We combine a recent approach for local low-rank approximation based on the Frobenius norm with a general empirical risk minimization for ranking losses.  ...  In this paper, we examine an alternative approach in which the rating matrix is locally low-rank.  ...  The improvement of LCR over GCR suggests that the locality assumption is more plausible and that the observed matrix is not well approximated by a single low-rank matrix.  ... 
doi:10.1145/2566486.2567970 dblp:conf/www/LeeBKLS14 fatcat:swyccgzerzf2hah363zatoyxey

A Fast Implementation of Singular Value Thresholding Algorithm using Recycling Rank Revealing Randomized Singular Value Decomposition [article]

Yaohang Li, Wenjian Yu
2017 arXiv   pre-print
A simulated annealing style cooling mechanism is employed to adaptively adjust the low-rank approximation precision threshold as SVT progresses.  ...  In this paper, we present a fast implementation of the Singular Value Thresholding (SVT) algorithm for matrix completion.  ...  for matrix completion is low-rank assumption, i.e., A has rank r m, n. Under the low-rank assumption, there exist an m × r matrix M and an r × n matrix N such that MN = A.  ... 
arXiv:1704.05528v1 fatcat:loovmbslbvgiffyew5xt5lo5b4

A Stochastic Sub-gradient Method for Low Rank Matrix Completion of Collaborative Recommendation

Weihua Yuan
2017 International Journal of Performability Engineering  
In this paper, we focus on nuclear norm regularized matrix completion model in large matrices, and propose a new model named stochastic sub-gradient method for low rank matrix completion (SS-LRMC).  ...  During iterations, we combine stochastic sub-gradient descent techniques with the adaptive singular value thresholding operator to obtain low rank intermediate solutions.  ...  Consequently, it is reasonable to assume the matrix low rank or approximately low rank and this is a typical matrix completion problem.  ... 
doi:10.23940/ijpe.17.05.p9.643656 fatcat:v4ivxsou6revvibr4b4t7l6jve

Predicting Missing Ratings in Recommender Systems: Adapted Factorization Approach

Carme Julià, Angel D. Sappa, Felipe Lumbreras, Joan Serrat, Antonio López
2009 International Journal of Electronic Commerce  
Experimental results with public data sets are provided to show that the proposed adapted factorization approach gives better predicted ratings than a widely used sVD-based approach.  ...  the paper presents a factorization-based approach to make predictions in recommender systems. these systems are widely used in electronic commerce to help customers find products according to their preferences  ...  The proposed adapted factorization approach gives the best low-rank matrix approximation to the data matrix.  ... 
doi:10.2753/jec1086-4415140203 fatcat:hfbbd73ihjdelfqujanfzb56z4

Algorithms and Literate Programs for Weighted Low-Rank Approximation with Missing Data [chapter]

Ivan Markovsky
2010 Springer Proceedings in Mathematics  
Linear models identification from data with missing values is posed as a weighted low-rank approximation problem with weights related to the missing values equal to zero.  ...  Alternating projections and variable projections methods for solving the resulting problem are outlined and implemented in a literate programming style, using Matlab/Octave's scripting language.  ...  I would like to thank Adam Prugel-Bennett and Mustansar Ghazanfar for discussions on the topic of recommender systems and for pointing out reference [17] and the MovieLens data set.  ... 
doi:10.1007/978-3-642-16876-5_12 fatcat:drnyo756jrbgpdilon2kuvyevu

ERMMA: Expected Risk Minimization for Matrix Approximation-based Recommender Systems

DongSheng Li, Chao Chen, Qin Lv, Li Shang, Stephen Chu, Hongyuan Zha
Matrix approximation (MA) is one of the most popular techniques in today's recommender systems.  ...  Based on the uniform stability theory, we propose an expected risk minimized matrix approximation method (ERMMA), which is designed to achieve better tradeoff between optimization error and generalization  ...  Recently, Srebro et al. (2004) analyzed the generalization error bounds of collaborative prediction with low-rank matrix approximation for binary recommendation problem. proposed the stable matrix approximation  ... 
doi:10.1609/aaai.v31i1.10743 fatcat:e665ljx7unberiti6tnzrahql4

Bilinear Generalized Approximate Message Passing—Part II: Applications

Jason T. Parker, Philip Schniter, Volkan Cevher
2014 IEEE Transactions on Signal Processing  
In this paper, we extend the generalized approximate message passing (G-AMP) approach, originally proposed for highdimensional generalized-linear regression in the context of compressive sensing, to the  ...  In Part I of this two-part paper, we derived our Bilinear G-AMP (BiG-AMP) algorithm as an approximation of the sum-product belief propagation algorithm in the high-dimensional limit, and proposed an adaptive  ...  We would also like to thank Subhojit Som, Jeremy Vila, and Justin Ziniel for helpful discussions about EM and turbo methods for AMP.  ... 
doi:10.1109/tsp.2014.2357773 fatcat:ptpfoi7hgrclrjz67jvnxn3rmu

Adaptive matrix completion for the users and the items in tail

Mohit Sharma, George Karypis
2019 The World Wide Web Conference on - WWW '19  
The low-rank matrix completion method is the state-of-the-art collaborative filtering method.  ...  Also, we show that the number of ratings that an item or a user has positively correlates with the ability of low-rank matrix-completion-based approaches to predict the ratings for the item or the user  ...  LLORMA assumes that the different parts of the user-item rating matrix can be approximated by different low-rank models and the complete user-item rating matrix is approximated as a weighted sum of these  ... 
doi:10.1145/3308558.3313736 dblp:conf/www/SharmaK19 fatcat:67eqbrvvcfhfnk73cqk2pw6yfi

Adaptively Weighted Top-N Recommendation for Organ Matching [article]

Parshin Shojaee, Xiaoyu Chen, Ran Jin
2021 arXiv   pre-print
AWTR sacrifices the overall recommendation accuracy by emphasizing the recommendation and ranking accuracy for top-N matched patients.  ...  In this paper, we formulate the organ matching decision-making as a top-N recommendation problem and propose an Adaptively Weighted Top-N Recommendation (AWTR) method.  ...  approximation of a matrix rank [7] .  ... 
arXiv:2107.10971v1 fatcat:hiw6hgjo3vax7pfxfq5hzpwxfq

Mixture-Rank Matrix Approximation for Collaborative Filtering

Dongsheng Li, Chao Chen, Wei Liu, Tun Lu, Ning Gu, Stephen M. Chu
2017 Neural Information Processing Systems  
Low-rank matrix approximation (LRMA) methods have achieved excellent accuracy among today's collaborative filtering (CF) methods.  ...  of the rating matrix, therefore leading to inferior recommendation accuracy.  ...  Thus, it is desirable to adopt mixture-rank matrix approximations rather than fixed-rank matrix approximations for recommendation tasks.  ... 
dblp:conf/nips/LiCLLGC17 fatcat:5zyytgqofnhahiqx2brmbx42de

Simple and effective neural-free soft-cluster embeddings for item cold-start recommendations

Shameem A. Puthiya Parambath, Sanjay Chawla
2020 Data mining and knowledge discovery  
Recommender systems are widely used in online platforms for easy exploration of personalized content.  ...  CIP can be used in conjunction with relevance ranking metrics like NDCG and MAP to measure the effectiveness of the cold-start recommendation algorithm.  ...  Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long  ... 
doi:10.1007/s10618-020-00708-6 fatcat:xhor425vmfe5bljgggn5bexpxi

Local Weighted Matrix Factorization for Top-n Recommendation with Implicit Feedback

Keqiang Wang, Hongwei Peng, Yuanyuan Jin, Chaofeng Sha, Xiaoling Wang
2016 Data Science and Engineering  
In this paper, we assume that the implicit feedback matrix has local property, where the original matrix is not globally low rank but some sub-matrices are low rank.  ...  In this paper, we propose Local Weighted Matrix Factorization (LWMF) for top-n recommendation by employing the kernel function to intensify local property and the weight function to model user preferences  ...  The original matrix is divided into several smaller sub-matrices, in which we can exploit local structures for better low-rank approximation.  ... 
doi:10.1007/s41019-017-0032-6 fatcat:hdbjtdp375ak7jp7mt2b3vxl74

CARec: Content-Aware Point-of-Interest Recommendation via Adaptive Bayesian Personalized Ranking

Baoping Liu, Yijun Su, Daren Zha, Neng Gao, Ji Xiang
2019 Australian Journal of Intelligent Information Processing Systems  
Then, by aggregating users' intrinsic preferences, we devise an adaptive Bayesian Personalized Ranking to generate the personalized ranked list of POIs for users.  ...  In this paper, we propose a novel content-aware POI recommendation framework via an adaptive Bayesian Personalized Ranking.  ...  Second, we make recommendations with an adaptive Bayesian Personalized Ranking model to achieve better performance.  ... 
dblp:journals/ajiips/LiuSZGX19 fatcat:cf2jqhefezfhfgdmderyoqhkbm

Robust Decentralized Low-Rank Matrix Decomposition

István Hegedűs, Árpád Berta, Levente Kocsis, András A. Benczúr, Márk Jelasity
2016 ACM Transactions on Intelligent Systems and Technology  
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.  ...  The local information at each node (personal attributes, documents, media ratings, etc.) defines one row in the matrix.  ...  Low rank matrix approximation can naturally be applied for collaborative filtering.  ... 
doi:10.1145/2854157 fatcat:wdctwhiuajenlcqyn7v4oedptu

Model reduction for transport-dominated problems via online adaptive bases and adaptive sampling [article]

Benjamin Peherstorfer
2020 arXiv   pre-print
Full-model solutions are approximated locally in time via local reduced spaces that are adapted with basis updates during time stepping.  ...  A core contribution of this work is an adaptive sampling scheme for selecting at which components to query the full model to compute basis updates.  ...  Acknowledgments The author would like to thank Nina Beranek (University of Ulm, Germany) for carefully reading an earlier version of this manuscript and for reporting typos and helpful comments.  ... 
arXiv:1812.02094v2 fatcat:mrrcdvahajgg3mrlvi2cljfwei
« Previous Showing results 1 — 15 out of 67,750 results