Filters








19 Hits in 2.7 sec

Diversifying Sparsity Using Variational Determinantal Point Processes [article]

Nematollah Kayhan Batmanghelich, Gerald Quon, Alex Kulesza, Manolis Kellis, Polina Golland, Luke Bornn
2014 arXiv   pre-print
We propose a novel diverse feature selection method based on determinantal point processes (DPPs).  ...  We introduce our approach in the context of Bayesian sparse regression, employing a DPP as a variational approximation to the true spike and slab posterior distribution.  ...  Determinantal Point Process The determinantal point process (DPP) defines distribution over configurations of points in space.  ... 
arXiv:1411.6307v1 fatcat:q5dhty7zwzcc3fqfsqkp7ze5pu

Repulsive Mixture Models of Exponential Family PCA for Clustering [article]

Maoying Qiao, Tongliang Liu, Jun Yu, Wei Bian, Dacheng Tao
2020 arXiv   pre-print
Specifically, a determinantal point process (DPP) is exploited as a diversity-encouraging prior distribution over the joint local EPCAs.  ...  An efficient variational EM algorithm is derived to perform parameter learning and hidden variable inference.  ...  DPP Determinantal point processes (DPP) are important statistical tools for diverse/repulsive relationship modelling.  ... 
arXiv:2004.03112v1 fatcat:f2dgmcwfyngh3kjuelpzhr7c6i

Exploring global diverse attention via pairwise temporal relation for video summarization [article]

Ping Li, Qinghao Ye, Luming Zhang, Li Yuan, Xianghua Xu, Ling Shao
2020 arXiv   pre-print
Most of existing systems employ encoder-decoder based recurrent neural networks, which fail to explicitly diversify the system-generated summary frames while requiring intensive computations.  ...  Particularly, determinantal point process results in different frame groups revealing diversified semantics in videos.  ...  [20] also developed DPP-LSTM, which further introduces Determinantal Point Process (DPP) to vsLSTM; Zhao et al.  ... 
arXiv:2009.10942v1 fatcat:pdv7i6ayurdrlieu5d5iu2gpvy

Hyperspectral Image Classification [chapter]

Rajesh Gogineni, Ashvini Chaturvedi
2019 Processing and Analysis of Hyperspectral Data [Working Title]  
Hyperspectral image (HSI) classification is a phenomenal mechanism to analyze diversified land cover in remotely sensed hyperspectral images.  ...  The determinantal point process (DPP) is used as a prior for probabilistic latent variable models in [68] . Probabilistic latent variable models are one of the vital elements of machine learning.  ...  The determinantal point process enables a modeler to specify a notion of similarity on the space of interest, which in this case is a space of possible latent distributions, via a positive definite kernel  ... 
doi:10.5772/intechopen.88925 fatcat:7ixv44bobbd3vkp7hn5c6tlb2y

Diversity in Machine Learning

Zhiqiang Gong, Ping Zhong, Weidong Hu
2019 IEEE Access  
Finally, we discuss some challenges of the diversity technology in machine learning and point out some directions in future work.  ...  Many factors can affect the performance of the machine learning process, among which the diversity of the machine learning process is an important one.  ...  the determinantal point process (DPP) measurement.  ... 
doi:10.1109/access.2019.2917620 fatcat:j7iam552ijhvvigzy5bdjwvttm

Diversified Bayesian Nonnegative Matrix Factorization

Qiao Maoying, Yu Jun, Liu Tongliang, Wang Xinchao, Tao Dacheng
2020 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
Specifically, the diversity prior is formulated with determinantal point processes (DPP) and is seamlessly embedded into a Bayesian NMF framework.  ...  In this paper, we approach this issue using a Bayesian framework.  ...  Determinantal Point Processes (DPP) Diversity is a good measure to subsets whenever the property of dissimilarity or repulsiveness is required.  ... 
doi:10.1609/aaai.v34i04.5991 fatcat:u2kkztzve5g5jpqrm72wy4e2ai

Testing Determinantal Point Processes [article]

Khashayar Gatmiry
2020 arXiv   pre-print
Determinantal point processes (DPPs) are popular probabilistic models of diversity. In this paper, we investigate DPPs from a new perspective: property testing of distributions.  ...  Practical diversified recommendations on YouTube with Determinantal Point Processes. In ACM International Conference on Information and Knowledge Management (CIKM), 2018.  ...  Fixed-point algorithms for learning determinantal point processes. In Int. Conference on Machine Learning (ICML), pages 2389-2397, 2015. [56] Jerzy Neyman and Egon Sharpe Pearson.  ... 
arXiv:2008.03650v1 fatcat:acj2dbojircenixcxoz3r2sifa

"Chitty-Chitty-Chat Bot": Deep Learning for Conversational AI

Rui Yan
2018 Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence  
It is useful to diversify the candidate response list using Maximum Marginal Relevance (MMR) or Determinantal Point Processes (DPP) .  ...  proposed to augment the Determinantal Point Processes (DPPs) with a Diversity Net so that the decoder selects items with good diversity and quality in balance.  ... 
doi:10.24963/ijcai.2018/778 dblp:conf/ijcai/Yan18 fatcat:s36y6mwhf5gapg3ila7nuh2fo4

Determinantal Point Processes for Machine Learning

Alex Kulesza
2012 Foundations and Trends® in Machine Learning  
Determinantal point processes (DPPs) are elegant probabilistic models of repulsion that arise in quantum physics and random matrix theory.  ...  This subset is distributed as a determinantal point process.  ...  Determinantal Point Processes Determinantal point processes (DPPs) were first identified as a class by Macchi [98] , who called them "fermion processes" because they give the distributions of fermion  ... 
doi:10.1561/2200000044 fatcat:qqq2fvo6tfcnbd73ee2oa43num

Diverse Counterfactual Explanations for Anomaly Detection in Time Series [article]

Deborah Sulem and Michele Donini and Muhammad Bilal Zafar and Francois-Xavier Aubet and Jan Gasthaus and Tim Januschowski and Sanjiv Das and Krishnaram Kenthapadi and Cedric Archambeau
2022 arXiv   pre-print
While in the previous methods, diversity is not explicitly enforced, the DiCE algorithm [23] includes a penalization on counterfactuals' similarity based on Determinantal Point Processes.  ...  We consider a general setting where time series are multivariate and the model processes all dimensions (or channels) jointly.  ... 
arXiv:2203.11103v1 fatcat:odumbeqcsvculm6jdakpyft6ei

Hyperspectral Image Classification – Traditional to Deep Models: A Survey for Future Prospects [article]

Muhammad Ahmad, Sidrah Shabbir, Swalpa Kumar Roy, Danfeng Hong, Xin Wu, Jing Yao, Adil Mehmood Khan, Manuel Mazzara, Salvatore Distefano, Jocelyn Chanussot
2021 arXiv   pre-print
Similarly, [192] combines multi-scale convolution-based CNN (MS-CNN) with diversified deep metrics based on determinantal point process (DPP) [193] priors for (1-D spectral, 2-D spectral-spatial, and  ...  However, this training process may result in two problems: first, multiple hidden units may tend to respond similarly [228] due to co-adaptation [229] and second is linked with the sparsity and selectivity  ... 
arXiv:2101.06116v2 fatcat:2duwvojkybgufo4kf6sbc6hdva

A Prior Guided Adversarial Representation Learning and Hypergraph Perceptual Network for Predicting Abnormal Connections of Alzheimer's Disease [article]

Qiankun Zuo, Baiying Lei, Shuqiang Wang, Yong Liu, Bingchuan Wang, Yanyan Shen
2021 arXiv   pre-print
Alzheimer's disease is characterized by alterations of the brain's structural and functional connectivity during its progressive degenerative processes.  ...  Concretely, a prior distribution from the anatomical knowledge is estimated to guide multimodal representation learning using an adversarial strategy.  ...  The Determinantal Point Process(DPP) [49] based prototype learning method is adopted to select a diversified prototype subset U 0 .  ... 
arXiv:2110.09302v1 fatcat:al4laaigjfbk3eotv6qnkbwxom

Orthogonality-Promoting Distance Metric Learning: Convex Relaxation and Theoretical Analysis [article]

Pengtao Xie, Wei Wu, Yichen Zhu, Eric P. Xing
2018 arXiv   pre-print
projection vectors in DML to be close to being orthogonal, to achieve three effects: (1) high balancedness -- achieving comparable performance on both frequent and infrequent classes; (2) high compactness -- using  ...  Determinantal Point Process [65] employed the determinantal point process (DPP) [27] as a prior to induce orthogonality in latent variable models.  ...  In these works, various orthogonality-promoting regularizers have been proposed, based on determinantal point process [27, 65] and cosine similarity [62, 3, 56] .  ... 
arXiv:1802.06014v1 fatcat:yrz7k5ck5nbulckuyhelufm4y4

Particle EM for Variable Selection

Veronika Ročková
2017 Journal of the American Statistical Association  
Motivated by non-parametric variational Bayes strategies, Particle EM achieves this goal by deploying an ensemble of interactive repulsive particles.  ...  This reconstruction reflects model selection uncertainty and is supported by asymptotic considerations, which indicate that the requisite number of particles need not be large in the presence of sparsity  ...  Acknowledgements The author would like to express gratitude to Ed George for providing useful feedback on earlier versions of this manuscript and to Gemma Moran for helping implement prototype versions  ... 
doi:10.1080/01621459.2017.1360778 fatcat:nrbc6jke6fcfvijeuygufn63oa

Diversity-Promoting and Large-Scale Machine Learning for Healthcare

Pengtao Xie
2019
In this network, a sequential LSTM is used to compose a sequence of tree LSTMs.  ...  Acknowledgments Related Works Larkey and Croft [211] studied the automatic assignment of ICD-9 codes to dictated inpatient discharge summaries, using a combination of three classifiers: k-nearest neighbors  ...  [16] investigated the "diversification" of Bayesian models using the determinantal point process (DPP) [204] prior. DPP has two drawbacks.  ... 
doi:10.1184/r1/7553468 fatcat:ac5ifp2lnzbk3hcupr2rszxj2m
« Previous Showing results 1 — 15 out of 19 results