Filters








5,665 Hits in 5.5 sec

Improving Query Quality for Transductive Learning in Learning to Rank

Xin Zhang, Zhi Cheng
2020 IEEE Access  
Of course, for scenarios involving no labeled data, the clustering-based transduction method is recommended for generating pseudo-labels for learning to rank. VOLUME XX, 2017 1 C.  ...  In this study, the authors presented a semi-supervised learning algorithm that used the clustered-based transductive method combined with a non-measurespecific listwise approach for learning to rank.   ... 
doi:10.1109/access.2020.3043459 fatcat:pt6w5g7dfvgapndct7prpznmfq

Generating Coherent Summaries with Textual Aspects

Renxian Zhang, Wenjie Li, Dehong Gao
2021 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
This paper presents a full-fledged system composed of three modules: finding sentence-level textual aspects, modeling aspect-based coherence with an HMM model, and selecting and ordering sentences with  ...  Due to limit of space, we report how our aspect-based HMM model compares with their aspect-agnostic model.  ...  Semi-supervised Learning We observe that for this task, classification accuracy may suffer from insufficient training data and a model learned from limited training data may not adapt well to unseen data  ... 
doi:10.1609/aaai.v26i1.8345 fatcat:rjqlli4xynfopeoxfoiopebm7i

Exploring knowledge of sub-domain in a multi-resolution bootstrapping framework for concept detection in news video

Gang Wang, Tat-Seng Chua, Ming Zhao
2008 Proceeding of the 16th ACM international conference on Multimedia - MM '08  
In this paper, we present a model based on a multi-resolution, multi-source and multi-modal (M3) bootstrapping framework that exploits knowledge of sub-domains for concept detection in news video.  ...  Because the characteristics and distributions of data in different sub-domains are different, we model and analyze the video in each sub-domain separately using a transductive framework.  ...  For example, Qi et al [16] proposed a transductive learning method to infer unlabeled test data by finding related labeled training data via a clustering method.  ... 
doi:10.1145/1459359.1459393 dblp:conf/mm/WangCZ08 fatcat:aeqh5tjd4rhr3nf4tqn75nwqwm

Analysis of the IJCNN 2011 UTL challenge

Isabelle Guyon, Gideon Dror, Vincent Lemaire, Daniel L. Silver, Graham Taylor, David W. Aha
2012 Neural Networks  
The second phase was dedicated to "crosstask transfer learning" (the competitors were provided with a limited amount of labeled data from "source tasks", distinct from the "target tasks").  ...  The goal was to learn data representations that capture regularities of an input space for re-use across tasks.  ...  This project is part of the DARPA Deep Learning program and is an activity of the Causality Workbench supported by the Pascal network of excellence funded by the European Commission and by the U.S.  ... 
doi:10.1016/j.neunet.2012.02.010 pmid:22374109 fatcat:gaoeexsksfaehaf5xaoirg773y

Graph Transduction Learning of Object Proposals for Video Object Segmentation [chapter]

Tinghuai Wang, Huiling Wang
2015 Lecture Notes in Computer Science  
Our core contribution is a graph transduction process that learns object proposals densely over space-time, exploiting both appearance models learned from rudimentary detections of sparse objectlike regions  ...  By learning a holistic model given a small set of objectlike regions, we propagate this prior knowledge of the recurring primary object to the rest of the video to generate a diverse set of object proposals  ...  All clusters are ranked based on the average score S(r) of its comprising regions.  ... 
doi:10.1007/978-3-319-16817-3_36 fatcat:mrzlxjps6rcwvau33gg23fkfh4

Semi-supervised ranking for document retrieval

Kevin Duh, Katrin Kirchhoff
2011 Computer Speech and Language  
Despite these successes, most algorithms to date construct ranking functions in a supervised learning setting, which assume that relevance labels are provided by human annotators prior to training the  ...  Recently there has been a surge of research in the field of "learning to rank", which aims at using labeled training data and machine learning algorithms to construct reliable ranking functions.  ...  Here we will call Problem One "Semi-supervised Rank Learning" and Problem Two "Learning to Rank with Missing Labels". See Figure 2 for a pictorial comparison.  ... 
doi:10.1016/j.csl.2010.05.002 fatcat:pccdrrrogfe5rjyfjgiogdinhi

Transductive Multilabel Learning via Label Set Propagation

Xiangnan Kong, Michael K. Ng, Zhi-Hua Zhou
2013 IEEE Transactions on Knowledge and Data Engineering  
In this paper, we study the problem of transductive multilabel learning and propose a novel solution, called TRAsductive Multilabel Classification (TRAM), to effectively assign a set of multiple labels  ...  We first formulate the transductive multilabel learning as an optimization problem of estimating label concept compositions.  ...  ACKNOWLEDGMENTS The authors wish to thank the editor and anonymous reviewers for their helpful comments and suggestions, and  ... 
doi:10.1109/tkde.2011.141 fatcat:csa2dtwitvczfcqcwcge33fn4a

Semi-supervised learning to rank with preference regularization

Martin Szummer, Emine Yilmaz
2011 Proceedings of the 20th ACM international conference on Information and knowledge management - CIKM '11  
We propose a semi-supervised learning to rank algorithm. It learns from both labeled data (pairwise preferences or absolute labels) and unlabeled data.  ...  Learning to Rank challenge 2010. The algorithm runs in linear time in the number of queries, and can work with huge datasets.  ...  Acknowledgements We thank Massih Reza Amini for making the semi-supervised RankBoost [2] implementation available, and Tom Minka and Vishwa Vinay for helpful discussions.  ... 
doi:10.1145/2063576.2063620 dblp:conf/cikm/SzummerY11 fatcat:yrr5hyxn2zel3jhjp3gbtnbyxa

Batch Mode Active Learning based on Multi-set Clustering

Yazhou Yang, Xiaoqing Yin, Yang Zhao, Jun Lei, Weili Li, Zhe Shu
2021 IEEE Access  
Our method utilizes a sequential active learner to retain the informativeness by providing a ranking of unlabeled samples and constructing multiple informative sets for the subsequent clusterings.  ...  Batch mode active learning, where a batch of samples is simultaneously selected and labeled, is a challenging task.  ...  Different from typical clustering-based approaches which use a fixed number of samples for clustering, ALMC first applies a clustering model on multiple informative sets and then adopts a selection criterion  ... 
doi:10.1109/access.2021.3053003 fatcat:h5nmakynrrd2bfbgfc55k5bfdy

Pool-Based Sequential Active Learning for Regression [article]

Dongrui Wu
2018 arXiv   pre-print
Active learning is a machine learning approach for reducing the data labeling effort.  ...  This paper focuses on pool-based sequential active learning for regression (ALR).  ...  It first builds a base model from a small number of labeled training samples, and then chooses a few most helpful unlabeled samples and queries for their labels.  ... 
arXiv:1805.04735v1 fatcat:3yws5o2vjvfzxnrnhrh3phkkjm

Effective and Efficient Data Poisoning in Semi-Supervised Learning [article]

Adriano Franci, Maxime Cordy, Martin Gubri, Mike Papadakis, Yves Le Traon
2020 arXiv   pre-print
Semi-Supervised Learning (SSL) aims to maximize the benefits of learning from a limited amount of labelled data together with a vast amount of unlabelled data.  ...  Moreover, our method can inform engineers of inputs that deserve investigation (relabelling them) before training the learning model.  ...  BACKGROUND AND RELATED WORK Semi-supervised learning [31] is a particular form of machine learning that attempts to maximize the benefits of learning from a limited amount of labelled data together with  ... 
arXiv:2012.07381v1 fatcat:owbnb7pnlzeb5gncploskfhbbe

Uncertainty sampling and transductive experimental design for active dual supervision

Vikas Sindhwani, Prem Melville, Richard D. Lawrence
2009 Proceedings of the 26th Annual International Conference on Machine Learning - ICML '09  
We apply classical uncertainty and experimental design based active learning schemes to graph/kernel-based dual supervision models.  ...  Empirical studies confirm the potential of these schemes to significantly reduce the cost of acquiring labeled data for training high-quality models.  ...  Acknowledgements We thank Yan Liu and Kai Yu for helpful conversations. We thank Gregory Druck for providing datasets for comparison with GE-FL.  ... 
doi:10.1145/1553374.1553496 dblp:conf/icml/SindhwaniML09 fatcat:iqdvno3c2vcxnnkeutfmkv6dzu

Intelligent Biometric Information Management

Harry Wechsler
2010 Intelligent Information Management  
We advance here a novel methodology for robust intelligent biometric information management with inferences and predictions made using randomness and complexity concepts.  ...  It further links, at the modeling and implementation level, the Bayesian framework, statistical learning theory (SLT) using transduction and semi-supervised learning, and Information Theory (IY) using  ...  An expanded framework that integrates graph-based semi-supervised learning and spectral clustering for the purpose of grouping and classification, i.e., label propagation, can be developed.  ... 
doi:10.4236/iim.2010.29060 fatcat:zp7qyfpwdzdo7dbsek2hkdgj2i

A Transductive Model-based Stress Recognition Method Using Peripheral Physiological Signals

Minjia Li, Lun Xie, Zhiliang Wang
2019 Sensors  
The proposed framework was presented as a transductive model based on transductive learning, which considered local learning as a virtue of the neighborhood knowledge of training examples.  ...  The degree of dispersion of the continuous labels in the y space was also one of the influencing factors of the transductive model.  ...  In order to address the above issues, a transductive model is proposed by combining transductive learning with clustering methods based on neighborhood knowledge in high dimension, as shown in Figure  ... 
doi:10.3390/s19020429 fatcat:onewqqphavbrtlx4wr4kqepxnq

Active Concept Learning in Image Databases

A. Dong, B. Bhanu
2005 IEEE Transactions on Systems Man and Cybernetics Part B (Cybernetics)  
To achieve concept learning, we a) propose a new user directed semi-supervised expectation-maximization algorithm for mixture parameter estimation, and b) develop a novel model selection method based on  ...  This paper presents an active concept learning approach based on the mixture model to deal with the two basic aspects of a database system: the changing (image insertion or removal) nature of a database  ...  For unsupervised learning, it is usually impossible to achieve satisfactory estimation for mixture model based on such limited number of samples in the high-dimensional feature space.  ... 
doi:10.1109/tsmcb.2005.846653 pmid:15971914 fatcat:prrpcpeqh5gbtfqpeecrsyjyti
« Previous Showing results 1 — 15 out of 5,665 results