Filters








384 Hits in 5.5 sec

A Tree-Based Method for Fast Repeated Sampling of Determinantal Point Processes

Jennifer Gillenwater, Alex Kulesza, Zelda Mariet, Sergei Vassilvitskii
2019 International Conference on Machine Learning  
It is often desirable in recommender systems and other information retrieval applications to provide diverse results, and determinantal point processes (DPPs) have become a popular way to capture the trade-off  ...  our approach runs over 300 times faster than traditional DPP sampling on collections of 100,000 items for samples of size 10.  ...  Determinantal point processes (DPPs) are probabilistic models of diverse subsets that have often been applied to these kinds of problems (Krause et al., 2008; Zhou et al., 2010; Lin & Bilmes, 2012; Chao  ... 
dblp:conf/icml/GillenwaterKMV19 fatcat:famaisrelrcgdi4szgj54bon5y

Dimensionality Reduction Based on Determinantal Point Process and Singular Spectrum Analysis for Hyperspectral Images

Zhijing Yang, Weizhao Chen, Faxian Cao, Yijun Yan, Meilin Wang, Chunmei Qing, Yongqiang Cheng
2018 IET Image Processing  
First, a new and fast band selection algorithm is proposed for hyperspectral images based on an improved determinantal point process (DPP).  ...  To reduce the amount of calculation, the dual-DPP is used for fast sampling representative pixels, followed by k-nearest neighbour-based local processing to explore more spatial information.  ...  Acknowledgments This work was supported by the National Nature Science Foundation of China (grant no. 61471132) and the Training program for outstanding young teachers in higher education institutions  ... 
doi:10.1049/iet-ipr.2018.5419 fatcat:rhecrda3azhfbnefy4c4nz4vie

Open Loop Hyperparameter Optimization and Determinantal Point Processes [article]

Jesse Dodge, Kevin Jamieson, Noah A. Smith
2019 arXiv   pre-print
In particular, we propose the use of k-determinantal point processes in hyperparameter optimization via random search.  ...  Driven by the need for parallelizable hyperparameter optimization methods, this paper studies open loop search methods: sequences that are predetermined and can be generated before a single configuration  ...  corner of the hypercube (as the true global optima are outside the space).  ... 
arXiv:1706.01566v4 fatcat:d5btomvhdvfdxp7da26ckujucq

Fixed-Size Determinantal Point Processes Sampling For Species Phylogeny

Diala Wehbe, Nicolas Wicker, Baydaa Al-Ayoubi, Luc Moulinier
2021 MathematicS in Action  
Determinantal point processes (DPPs) are popular tools that supply useful information for repulsiveness.  ...  The tree sampling task is important in many studies in modern bioinformatics. The results show a fast mixing sampler for k-DPP, for which a polynomial bound on the mixing time is given.  ...  Thompson for careful reading of the manuscript and English language correction.  ... 
doi:10.5802/msia.13 fatcat:lezkimjydnfvneywnjpg3lpel4

Large-data determinantal clustering [article]

Serge Vicente, Alejandro Murua
2021 arXiv   pre-print
Based on a determinantal point process or DPP sampling, it ensures that subsets of similar points are less likely to be selected as centroids. It favors more diverse subsets of points.  ...  The sampling algorithm of the determinantal point process requires the eigendecomposition of a Gram matrix. This becomes computationally intensive when the data size is very large.  ...  Determinantal point processes, or DPPs for short, introduced by [8] , can address this problem.  ... 
arXiv:2102.03954v1 fatcat:umw5ergxtna5lfsuduhfunnyh4

In Conclusion Not Repetition: Comprehensive Abstractive Summarization With Diversified Attention Based On Determinantal Point Processes [article]

Lei Li, Wei Liu, Marina Litvak, Natalia Vanetik, Zuying Huang
2019 arXiv   pre-print
We introduce Diverse Convolutional Seq2Seq Model(DivCNN Seq2Seq) using Determinantal Point Processes methods(Micro DPPs and Macro DPPs) to produce attention distribution considering both quality and diversity  ...  Without breaking the end to end architecture, DivCNN Seq2Seq achieves a higher level of comprehensiveness compared to vanilla models and strong baselines.  ...  In this paper we propose a novel Diverse Convolutional Seq2Seq Model(DivCNN Seq2Seq) based on Micro Determinantal Point Processes(Micro DPPs) and Macro Determinantal Point Processes(Macro DPPs).  ... 
arXiv:1909.10852v1 fatcat:xnuua3m36ndfhdxpjhp7oludzi

Scalable MCMC Sampling for Nonsymmetric Determinantal Point Processes [article]

Insu Han, Mike Gartrell, Elvis Dohmatob, Amin Karbasi
2022 arXiv   pre-print
A determinantal point process (DPP) is an elegant model that assigns a probability to every subset of a collection of n items.  ...  Our method is based on a state-of-the-art NDPP rejection sampling algorithm, which we enhance with a novel approach for efficiently constructing the proposal distribution.  ...  Amin Karbasi acknowledges funding in direct support of this work from NSF (IIS-1845032), ONR (N00014-19-1-2406), and the AI Institute for Learning-Enabled Optimization at Scale (TILOS).  ... 
arXiv:2207.00486v1 fatcat:4wdfyck74rfmxf3dqcfklibb7q

Sampling Unknown Decision Functions to Build Classifier Copies [article]

Irene Unceta, Diego Palacios, Jordi Nin, Oriol Pujol
2019 arXiv   pre-print
A fundamental step of the copying process is generating an unlabelled set of points to explore the decision behavior of the targeted classifier throughout the input space.  ...  In this article we propose two sampling strategies to produce such sets. We validate them in six well-known problems and compare them with two standard methods.  ...  As a future work we will investigate how to make Bayesian sampling faster and search new sampling methods, such as Determinantal Point Processes (DPP).  ... 
arXiv:1910.00237v1 fatcat:5qvh7jve6jgnhc5jtdounmoryu

MAP Inference for Customized Determinantal Point Processes via Maximum Inner Product Search

Insu Han, Jennifer Gillenwater
2020 International Conference on Artificial Intelligence and Statistics  
Determinantal point processes (DPPs) are a good fit for modeling diversity in many machine learning applications.  ...  For instance, in recommender systems, one might have a basic DPP defined by item features, and a customized version of this DPP for each user with features re-weighted according to user preferences.  ...  We also use L A as shorthand for the square matrix L AA , and colon to indicate all rows or columns: L A: ∈ R |A|×N . 2.1 Determinantal point processes (DPPs) A DPP on a set of N items defines a probability  ... 
dblp:conf/aistats/HanG20 fatcat:qded5owlz5bqjniatwfp46gg6q

Diversity-Aware Batch Active Learning for Dependency Parsing [article]

Tianze Shi, Adrian Benton, Igor Malioutov, Ozan İrsoy
2021 arXiv   pre-print
In particular, we investigate whether enforcing diversity in the sampled batches, using determinantal point processes (DPPs), can improve over their diversity-agnostic counterparts.  ...  Additionally, our diversityaware strategy is robust under a corpus duplication setting, where diversity-agnostic sampling strategies exhibit significant degradation.  ...  Acknowledgements We thank the anonymous reviewers for their insightful reviews, and Prabhanjan Kambadur, Chen-Tse Tsai, and Minjie Xu for discussion and comments.  ... 
arXiv:2104.13936v1 fatcat:35dwwvledzairmvtbw6wf5shqi

Scalable Sampling for Nonsymmetric Determinantal Point Processes [article]

Insu Han, Mike Gartrell, Jennifer Gillenwater, Elvis Dohmatob, Amin Karbasi
2022 arXiv   pre-print
A determinantal point process (DPP) on a collection of M items is a model, parameterized by a symmetric kernel matrix, that assigns a probability to every subset of those items.  ...  In our experiments we compare the speed of all of these samplers for a variety of real-world tasks.  ...  All of the code implementing our constrained learning and sampling algorithms is publicly available † † . The proofs for our theoretical contributions are available in Appendix E.  ... 
arXiv:2201.08417v2 fatcat:rskpiwmvz5b4xge2l4qhsizxmm

Recent Advances in Diversified Recommendation [article]

Qiong Wu, Yong Liu, Chunyan Miao, Yin Zhao, Lu Guan, Haihong Tang
2019 arXiv   pre-print
With the rapid development of recommender systems, accuracy is no longer the only golden criterion for evaluating whether the recommendation results are satisfying or not.  ...  Specifically, we first review the various definitions of diversity and generate a taxonomy to shed light on how diversity have been modeled or measured in recommender systems.  ...  With such data, there are three major approaches for trading-off between relevance and diversity, i.e., post-processing methods, learning-to-rank methods, and determinantal point process based methods.  ... 
arXiv:1905.06589v1 fatcat:yzzea2ozkre67bt646ic4vij6u

Approximate Bayesian Computation and Model Validation for Repulsive Spatial Point Processes [article]

Shinichiro Shirota, Alan. E. Gelfand
2016 arXiv   pre-print
A recent alternative, at least to the statistical community, is the determinantal point process.  ...  We provide details for all of the above along with some simulation investigation and an illustrative analysis of a point pattern of tree data exhibiting repulsion.  ...  Acknowledgements The work of the first author was supported in part by the Nakajima Foundation.  ... 
arXiv:1604.07027v2 fatcat:67j6mkrz7nfjzmwxtiooetg57e

Large-Margin Determinantal Point Processes [article]

Boqing Gong, Wei-lun Chao, Kristen Grauman, Fei Sha
2014 arXiv   pre-print
Determinantal point processes (DPPs) offer a powerful approach to modeling diversity in many applications where the goal is to select a diverse subset.  ...  Second, we propose a novel parameter estimation technique based on the principle of large margin separation.  ...  Background: Determinantal point processes We first review background on the determinantal point process (DPP) [11] and the standard maximum likelihood estimation technique for learning DPP parameters  ... 
arXiv:1411.1537v2 fatcat:udac2uacevdojkqz5ox4b2gq24

Special Issue of Journal of Statistical Physics Devoted to Complex Networks

Diego Garlaschelli, Remco van der Hofstad, Frank den Hollander, Michel Mandjes
2018 Journal of statistical physics  
It is a crossroad of concepts, ideas and techniques that bring together scientists from different disciplines, all facing the major challenges that arise from dealing with the measurement, simulation,  ...  modelling and analysis of (typically very large) real-world net-B Frank den Hollander  ...  like the massive Gaussian free field and determinantal point processes with fast decaying kernels.  ... 
doi:10.1007/s10955-018-2166-y fatcat:nsqkugzycbbr5prm3fknmvd5b4
« Previous Showing results 1 — 15 out of 384 results