3,020 Hits in 5.1 sec

Conditional Bernoulli Mixtures for Multi-label Classification

Cheng Li, Bingyu Wang, Virgil Pavlu, Javed A. Aslam
2016 International Conference on Machine Learning  
In this paper, we propose a new multi-label classification method based on Conditional Bernoulli Mixtures.  ...  Multi-label classification is an important machine learning task wherein one assigns a subset of candidate labels to an object.  ...  Acknowledgements We thank reviewers for their helpful comments, Shivani Agarwal for her suggestions on related work, and Pavel Metrikov for discussions on the model.  ... 
dblp:conf/icml/LiWPA16 fatcat:lrmr56w3z5eudfqyfvcnj5yo6q

Dirichlet-Bernoulli Alignment: A Generative Model for Multi-Class Multi-Label Multi-Instance Corpora

Shuang-Hong Yang, Hongyuan Zha, Bao-Gang Hu
2009 Neural Information Processing Systems  
By casting predefined classes as latent Dirichlet variables (i.e., instance level labels), and modeling the multi-label of each pattern as Bernoulli variables conditioned on the weighted empirical average  ...  DBA is useful for both pattern classification and instance disambiguation, which are tested on text classification and named entity disambiguation in web search queries respectively.  ...  from a Bernoulli distribution conditioned on all the sampled labels used for generating its instances.  ... 
dblp:conf/nips/YangZH09 fatcat:xpii7xbkb5dozf2eib545iryv4

Scalable Text Classification with Sparse Generative Modeling [chapter]

Antti Puurula
2012 Lecture Notes in Computer Science  
We report on classification experiments on 5 publicly available datasets for large-scale multi-label classification.  ...  We reduce the time complexity of Multinomial Naive Bayes classification with sparsity and show how to extend these findings into three multi-label extensions: Binary Relevance, Label Powerset and Multi-label  ...  Multi-label Mixture Model Multi-label mixture models [12, 13, 14] attempt generalization of MNB to multilabel data by decomposing labelset-conditional Multinomials into mixtures of label-conditional  ... 
doi:10.1007/978-3-642-32695-0_41 fatcat:zi6k53dt4fbv3g4xpzv7djx3dq

Classification of weak multi-view signals by sharing factors in a mixture of Bayesian group factor analyzers [article]

Sami Remes and Tommi Mononen and Samuel Kaski
2016 arXiv   pre-print
We propose a novel classification model for weak signal data, building upon a recent model for Bayesian multi-view learning, Group Factor Analysis (GFA).  ...  Motivation for the setting comes from single-trial functional brain imaging data, having a very low signal-to-noise ratio and a natural multi-view setting, with the different sensors, measurement modalities  ...  Zhdanov for the possibility to use the anonymized dual-MEG data, A. Mandel for preprocessing of the data set, and L. Hirvenkari for the stimulus timing.  ... 
arXiv:1512.05610v2 fatcat:rzbwbgtucncgbhsgadivwcyfvu

Relaxed Multivariate Bernoulli Distribution and Its Applications to Deep Generative Models

Xi Wang, Junming Yin
2020 Conference on Uncertainty in Artificial Intelligence  
We demonstrate its effectiveness in two tasks: density estimation with Bernoulli VAE and semisupervised multi-label classification.  ...  In this work, we propose a multivariate generalization of the Relaxed Bernoulli distribution, which can be reparameterized and can capture the correlation between variables via a Gaussian copula.  ...  Acknowledgements We thank the reviewers for their constructive feedback. This work is partially supported by Amazon AWS Machine Learning Research Award (JY).  ... 
dblp:conf/uai/WangY20 fatcat:sh7wpgilynd2pngyw3ub4m3nxi

On Word Frequency Information and Negative Evidence in Naive Bayes Text Classification [chapter]

Karl-Michael Schneider
2004 Lecture Notes in Computer Science  
One version, called multi-variate Bernoulli or binary independence model, uses binary word occurrence vectors, while the multinomial model uses word frequency counts.  ...  Many publications cite this difference as the main reason for the superior performance of the multinomial Naive Bayes classifier. We argue that this is not true.  ...  Spam recall for multi-variate Bernoulli and multinomial Naive Bayes on the ling-spam corpus, with 10-fold cross validation.  ... 
doi:10.1007/978-3-540-30228-5_42 fatcat:u7zr7ip3fjckjd4kocorlnnrf4

Distributed Fusion With Multi-Bernoulli Filter Based on Generalized Covariance Intersection

Bailu Wang, Wei Yi, Reza Hoseinnezhad, Suqi Li, Lingjiang Kong, Xiaobo Yang
2017 IEEE Transactions on Signal Processing  
To solve this challenging problem, we firstly approximate the fused posterior as the unlabeled version of δ-generalized labeled multi-Bernoulli (δ-GLMB) distribution, referred to as generalized multi-Bernoulli  ...  In this paper, we propose a distributed multi-object tracking algorithm through the use of multi-Bernoulli (MB) filter based on generalized Covariance Intersection (G-CI).  ...  Labeled Multi Bernoulli (δ-GLMB) distribution is defined for labeled RFSs.  ... 
doi:10.1109/tsp.2016.2617825 fatcat:5qtt5pix3beb3bwdd2udhcwn7i

Multi-target joint detection, tracking and classification based on random finite set for aerospace applications

Zhongliang Jing, Minzhe Li, Henry Leung
2018 Aerospace Systems  
It is observed that the labeled random finite set theory provides a good foundation for a joint solution for multi-target detection, tracking and classification.  ...  This paper summarized the existing work for multi-target joint detection, tracking and classification based on labeled random finite set.  ...  Based on the labeled RFS, the GLMB filter for multi-target tracking can be derived as follows: Assume that the multi-target prior distribution is the labeled multi-Bernoulli distribution on the state space  ... 
doi:10.1007/s42401-018-0003-2 fatcat:e7zbzdjhfvgfjf2cttvnqtcyl4

Multi-target Joint Detection, Tracking and Classification Based on Generalized Bayesian Risk using Radar and ESM sensors [article]

Minzhe Li, Zhongliang Jing
2018 arXiv   pre-print
Furthermore, the conditional labeled multi-Bernoulli filter is developed to calculate the estimates and costs for different hypotheses and decisions of target classes using attribute and dynamical measurements  ...  In this paper, a novel approach is proposed for multi-target joint detection, tracking and classification based on the labeled random finite set and generalized Bayesian risk using Radar and ESM sensors  ...  Labeled multi-Bernoulli RFS and multi-target Bayes filter In [10] , a Bernoulli RFS was used to represents the uncertainty about the existence of a single object.  ... 
arXiv:1807.02267v1 fatcat:znow46qul5epvfctedjdnu3rfa

Multi-Label Annotation Aggregation in Crowdsourcing [article]

Xuan Wei, Daniel Dajun Zeng, Junming Yin
2020 arXiv   pre-print
In this paper, we present new flexible Bayesian models and efficient inference algorithms for multi-label annotation aggregation by taking both annotator reliability and label dependency into account.  ...  One of the obvious challenges is how to aggregate these possibly noisy labels provided by a set of heterogeneous annotators.  ...  Mixture of multiple Bernoulli distribution has been utilized to model label dependency for supervised multi-label classification task; see [17] and references therein.  ... 
arXiv:1706.06120v2 fatcat:m5hpqoyyinf7xakpwm2n4yedyu

A New Fine-Grained Weighting Method in Multi-Label Text Classification

Chang-Hwan Lee
2014 Midwest Artificial Intelligence and Cognitive Science Conference  
Multi-label classification is one of the important research areas in data mining. In this paper, a new multilabel classification method using multinomial naive Bayes is proposed.  ...  We use a new fine-grained weighting method for calculating the weights of feature values in multinomial naive Bayes.  ...  Based on this model a multi-label document is produced by a mixture of the word distributions of its labels.  ... 
dblp:conf/maics/Lee14 fatcat:jzznjj2lnfdojl6vri7ppth3ae

Hetero-Labeled LDA: A Partially Supervised Topic Model with Heterogeneous Labels [chapter]

Dongyeop Kang, Youngja Park, Suresh N. Chari
2014 Lecture Notes in Computer Science  
accommodate labels for only a subset of classes (i.e., partial labels).  ...  resulting in better classification and clustering accuracy than existing supervised or semisupervised topic models.  ...  Also, to further improve the performance of label prediction for partially labeled documents, we consider generating topic hierarchies such as Hierarchical Dirichlet Process (HDP) [23] .  ... 
doi:10.1007/978-3-662-44848-9_41 fatcat:sv4tdozw6narpgdjzna7osdjvq

Scalable Generative Models for Multi-label Learning with Missing Labels

Vikas Jain, Nirbhay Modhe, Piyush Rai
2017 International Conference on Machine Learning  
We present a scalable, generative framework for multi-label learning with missing labels.  ...  Our framework consists of a latent factor model for the binary label matrix, which is coupled with an exposure model to account for label missingness (i.e., whether a zero in the label matrix is indeed  ...  Incorporating these latent variables leads to improve multi-label classification accuracies, and also enables doing interesting qualitative analyses.  ... 
dblp:conf/icml/JainMR17 fatcat:s5fn3uswlres7monrx5o4jhbzi

Label Correlation Mixture Model: A Supervised Generative Approach to Multilabel Spoken Document Categorization

Zhiyang He, Ji Wu, Tao Li
2015 IEEE Transactions on Emerging Topics in Computing  
The multilabel conditioned document model can be regarded as a supervised label mixture model, in which labels for a document are known. Each label is characterized by distributions over words.  ...  For the parameter learning of the multilabel conditioned document model, in addition to maximumlikelihood estimation, a discriminative approach based on the minimum classification error rate training is  ...  A mixture model approach was proposed in [9] for multi-label text classification, in which each label was regarded as a class and modeled by a word distribution.  ... 
doi:10.1109/tetc.2014.2377559 fatcat:d6gcguctpjgfnmwivjwlevfgim

Generative Models for Labeling Multi-object Configurations in Images [chapter]

Yali Amit, Alain Trouvé
2006 Lecture Notes in Computer Science  
Training involves estimation of model parameters for each class separately. Both training and classification involve estimation of hidden pose variables which can be computationally intensive.  ...  The models assume conditional independence of binary oriented edge variables conditional on a hidden instantiation parameter, which also determines an object support.  ...  If we put aside the approach of bottom up segmentation and subsequent classification, we need to be able to combine detection and classification for multi-object configurations.  ... 
doi:10.1007/11957959_19 fatcat:wd4oshqdircrlcbxtcffsgtxiq
« Previous Showing results 1 — 15 out of 3,020 results