1,387 Hits in 5.3 sec

Instance- and bag-level manifold regularization for aggregate outputs classification

Shuo Chen, Bin Liu, Mingjie Qian, Changshui Zhang
2009 Proceeding of the 18th ACM conference on Information and knowledge management - CIKM '09  
The framework can be of both instance level and bag level for different testing goals.  ...  We focus on the aggregate outputs classification problem in this paper, and set up a manifold regularization framework to deal with it.  ...  CONCLUSION In this paper, we set up a manifold regularization framework for the AOC problem, and propose four new algorithms for both instance-and bag-level settings.  ... 
doi:10.1145/1645953.1646180 dblp:conf/cikm/ChenLQZ09 fatcat:3djlgij67vhtxmstmgisputih4

Learning from Label Proportions with Consistency Regularization [article]

Kuen-Han Tsai, Hsuan-Tien Lin
2019 arXiv   pre-print
The problem of learning from label proportions (LLP) involves training classifiers with weak labels on bags of instances, rather than strong labels on individual instances.  ...  the data manifold.  ...  Acknowledgments We would like to acknowledge Michelle Yuan, Si-An Chen, Chien-Min Yu for fruitful discussion and valuable suggestions that helped to improve this article.  ... 
arXiv:1910.13188v1 fatcat:67zg7sitsfdsxohhfejasixtmm

Variational Learning on Aggregate Outputs with Gaussian Processes [article]

Ho Chung Leon Law, Dino Sejdinovic, Ewan Cameron, Tim CD Lucas, Seth Flaxman, Katherine Battle, Kenji Fukumizu
2018 arXiv   pre-print
to outputs at a much coarser level than that of the inputs.  ...  We consider an approach to this problem based on variational learning with a model of output aggregation and Gaussian processes, where aggregation leads to intractability of the standard evidence lower  ...  Acknowledgement We thank Kaspar Martens for useful discussions, and Dougal Sutherland for providing the code base in which this work was based on.  ... 
arXiv:1805.08463v1 fatcat:p6f36i2ipjdvffnp75yxtaa3om

A Multiple Instance Learning Framework for Identifying Key Sentences and Detecting Events

Wei Wang, Yue Ning, Huzefa Rangwala, Naren Ramakrishnan
2016 Proceedings of the 25th ACM International on Conference on Information and Knowledge Management - CIKM '16  
Using a multiple instance learning approach, we take advantage of the fact that while labels at the sentence level are difficult to obtain, they are relatively easy to gather at the document level.  ...  Our model, trained without annotated sentence labels, yields performance that is competitive with selected state-of-the-art models for event detection and sentence identification.  ...  this work for Governmental purposes notwithstanding any copyright annotation thereon.  ... 
doi:10.1145/2983323.2983821 dblp:conf/cikm/0064NRR16 fatcat:q64ozniftbbbrmil4ekohj3xya

From Group to Individual Labels Using Deep Features

Dimitrios Kotzias, Misha Denil, Nando de Freitas, Padhraic Smyth
2015 Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining - KDD '15  
One context in which this occurs is where we have labels for groups of instances but not for the instances themselves, as in multi-instance learning.  ...  In many classification problems labels are relatively scarce.  ...  ACKNOWLEDGEMENTS The authors would like to thank Eric Nalisnick, Jihyun Park, Kevin Bache and Moshe Lichman for their help in the manual tagging of sentences and fruitful discussions over the course of  ... 
doi:10.1145/2783258.2783380 dblp:conf/kdd/KotziasDFS15 fatcat:klazq7eph5avhbqdq3xwpxsehm

Deep Learning of Segment-Level Feature Representation with Multiple Instance Learning for Utterance-Level Speech Emotion Recognition

Shuiyang Mao, P.C. Ching, Tan Lee
2019 Interspeech 2019  
For the segment-level classification, we attempt two different deep neural network (DNN) architectures called SegMLP and SegCNN, respectively.  ...  Then the utterance-level classification is constructed as an aggregation of the segment-level decisions.  ...  In MIL, the training set contains labeled bags that are comprised of many unlabeled instances, and the task is to predict the labels of unseen bags and instances.  ... 
doi:10.21437/interspeech.2019-1968 dblp:conf/interspeech/MaoCL19 fatcat:tokmuji5vbaxloskv5nc46f3bq

From Language to Location Using Multiple Instance Neural Networks [chapter]

Sneha Nagpaul, Huzefa Rangwala
2018 Lecture Notes in Computer Science  
Bag level model: N is set to 10 and the outputs of the instance level models are averaged at a higher layer for the bag level output.  ...  judging is presented in [7] , where instances are triaged and classification on a bag level is subsequently evaluated.  ... 
doi:10.1007/978-3-319-93372-6_5 fatcat:tnnkeg4rf5cxtigkm5z66yyujm

Short Text Embedding Autoencoders with Attention-based Neighborhood Preservation

Chao Wei, Lijun Zhu, Jiaoxiang Shi
2020 IEEE Access  
The Random Forest (RF) is an exemplar of the ensemble learning method for classification, which combines the random subspace method and bagging method.  ...  of self-reconstruction and manifold graph regularization.  ... 
doi:10.1109/access.2020.3042778 fatcat:k4rn6rwurzcpheet74zudnfwfu

Context-Aware Multi-instance Learning Based on Hierarchical Sparse Representation

Bing Li, Weihua Xiong, Weiming Hu
2011 2011 IEEE 11th International Conference on Data Mining  
More recently, researchers focus on two important issues for MIL: Instances' contextual structures representation in the same bag and online MIL schemes.  ...  We firstly construct the inner contextual structure among instances in the same bag based on a novel sparse -graph.  ...  Algorithm 1 sparse -graph construction for each bag. 1: Input: A bag in MIL as = { ,1 , ,2 , ..., , } ⊆ , regularization coefficient and locality threshold 2: For = 1 : Table II ONLINE II UPDATE FOR  ... 
doi:10.1109/icdm.2011.43 dblp:conf/icdm/LiXH11 fatcat:2lae6alw65hnfcw4aprqjp7xai

Deep Multi-Instance Transfer Learning [article]

Dimitrios Kotzias, Misha Denil, Phil Blunsom, Nando de Freitas
2014 arXiv   pre-print
This approach, which combines ideas from transfer learning, deep learning and multi-instance learning, reduces the need for laborious human labelling of fine-grained data when abundant labels are available  ...  at the group level.  ...  Using these features, we formulate a regularized manifold learning objective function to learn the labels of each sentence.  ... 
arXiv:1411.3128v2 fatcat:5ghubtop25cprj376kqhxxwi7e

Learning to Aggregate Using Uninorms [chapter]

Vitalik Melnikov, Eyke Hüllermeier
2016 Lecture Notes in Computer Science  
Experimental results for a corresponding model are presented for a review data set, for which the aggregation problem consists of combining different reviewer opinions about a paper into an overall decision  ...  In this paper, we propose a framework for a class of learning problems that we refer to as "learning to aggregate".  ...  Acknowledgments We thank Pritha Gupta and Karlson Pfannschmidt for their helpful suggestions.  ... 
doi:10.1007/978-3-319-46227-1_47 fatcat:ngvwlp22aza6fihtb6oodbpj6q

Not-so-supervised: a survey of semi-supervised, multi-instance, and transfer learning in medical image analysis [article]

Veronika Cheplygina, Marleen de Bruijne, Josien P. W. Pluim
2018 arXiv   pre-print
We also discuss connections between these learning scenarios, and opportunities for future research.  ...  We review semi-supervised, multiple instance, and transfer learning in medical imaging, both in diagnosis/detection or segmentation tasks.  ...  Because instance-level and some bag-level classifiers can provide instance labels, the focus of MIL became two-fold: classifying bags and classifying instances.  ... 
arXiv:1804.06353v2 fatcat:xke66fsrgjanjmopwti5tze4fm

Not-so-supervised: a survey of semi-supervised, multi-instance, and transfer learning in medical image analysis

Veronika Cheplygina, Marleen de Bruijne, Josien P.W. Pluim
2019 Medical Image Analysis  
We also discuss connections between these learning scenarios, and opportunities for future research.  ...  We give an overview of semi-supervised, multiple instance, and transfer learning in medical imaging, both in diagnosis or segmentation tasks.  ...  Ilse, Ragav Venkatesan and Wouter Kouw.  ... 
doi:10.1016/ pmid:30959445 fatcat:bbgz7v3ixvggnksuvxsxernobm

Deep Multiple Instance Feature Learning via Variational Autoencoder [article]

Shabnam Ghaffarzadegan
2018 arXiv   pre-print
MIL is a weakly supervised learning problem where labels are associated with groups of instances (referred as bags) instead of individual instances.  ...  differences between latent representations of all instances and negative instances.  ...  [1] proposed two different methods namely mi-SVM and MI-SVM, for instance-level and baglevel classification respectively, to define the margin for positive bags.  ... 
arXiv:1807.02490v1 fatcat:fkbqveoz5fdlrixaizssgben5q

Semi-supervised multi-instance multi-label learning for video annotation task

Xin-Shun Xu, Yuan Jiang, Xiangyang Xue, Zhi-Hua Zhou
2012 Proceedings of the 20th ACM international conference on Multimedia - MM '12  
This approach takes label correlations into account, and enforces similar instances to share similar multi-labels.  ...  Indeed, the video annotation task is inherently a Multi-Instance Multi-Label learning (MIML) problem.  ...  For ML-GRF, we aggregate all instances in a bag into one feature vector; for MISSL, we run it for multiple times each for one label; for MIML-SVM, we simply neglect unlabeled data.  ... 
doi:10.1145/2393347.2396300 dblp:conf/mm/XuJXZ12 fatcat:sjrz2obs2fbcte6uafxovtdlmi
« Previous Showing results 1 — 15 out of 1,387 results