Filters








10,136 Hits in 4.3 sec

Multi-view learning from imperfect tagging

Zhongang Qi, Ming Yang, Zhongfei (Mark) Zhang, Zhengyou Zhang
2012 Proceedings of the 20th ACM international conference on Multimedia - MM '12  
The main idea of MITL lies in extracting the information of the imperfectly tagged training dataset from multiple views to differentiate the data points in the role of classification.  ...  The proposed methods can not only complete the incomplete tagging but also denoise the noisy tagging through an inductive learning.  ...  The F 1 a \F 1 i for the testing set in two views combined on incomplete tagging learning and on noisy data learning, respectively.  ... 
doi:10.1145/2393347.2393416 dblp:conf/mm/QiYZZ12 fatcat:2weujwshnbekhkexptquep3lmy

A Transfer-Based Additive LS-SVM Classifier for Handling Missing Data

Guanjin Wang, Jie Lu, Kup-Sze Choi, Guangquan Zhang
2018 IEEE Transactions on Cybernetics  
can be used to enhance the classification performance on incomplete training datasets.  ...  This method also simultaneously determines the influence of classification errors caused by each incomplete sample using a fast leave-one-out cross validation strategy, as an alternative way to clean the  ...  Methods in the third category estimate the data distributions of the complete and incomplete data portions in the dataset , and make use of them for pattern classification.  ... 
doi:10.1109/tcyb.2018.2872800 pmid:30334775 fatcat:vcavjdyeyng5llxtgzak5qvqoi

Using Support Vector Machines for Classifying Large Sets of Multi-Represented Objects [chapter]

Hans-Peter Kriegel, Peer Kröger, Alexey Pryakhin, Matthias Schubert
2004 Proceedings of the 2004 SIAM International Conference on Data Mining  
Databases are a key technology for molecular biology which is a very data intensive discipline.  ...  Therefore, our method introduces the technique of objectadjusted weighting which regulates the impact of each representation dynamically for each object.  ...  In the case of Set 5 the classification accuracy of 73.41% even exceeded the values observed for sequences only (71.09%). Thus, the system is able to handle incomplete data.  ... 
doi:10.1137/1.9781611972740.10 dblp:conf/sdm/KriegelKPS04 fatcat:z5busyjmj5dcpgxigre3u3isui

Web Page Structure Enhanced Feature Selection for Classification of Web Pages

B. LeelaDevi, A. Sankar
2013 International Journal of Computer Applications  
Semantic search motivates Semantic Web from inception for classification and retrieval processes.  ...  The features are classified using Support Vector Machine (SVM) using different kernels.  ...  SVMs are a useful tool for insolvency analysis, in case of data non-regularity, for example when data is not regularly distributed/having an unknown distribution [19] .  ... 
doi:10.5120/11818-7494 fatcat:ol7zrmg645fpjp3mv3uhu7tbie

Web image gathering with region-based bag-of-features and multiple instance learning

Keiji Yanai
2009 2009 IEEE International Conference on Multimedia and Expo  
By this region-based classification, we can separate foreground regions from background regions and achieve more effective image training from incomplete training data.  ...  The contribution of this work is introducing the region-based bag-of-features representation into an Web image gathering task where training data is incomplete and having proved its effectiveness by comparing  ...  The contribution of this work is introducing the region-based bag-of-features representation and multipleinstance SVM into an Web image gathering task where training data is incomplete.  ... 
doi:10.1109/icme.2009.5202531 dblp:conf/icmcs/Yanai09 fatcat:g5zs22nd25da3hoyl422dz5rwm

On Multilabel Classification Methods of Incompletely Labeled Biomedical Text Data

Anton Kolesov, Dmitry Kamyshenkov, Maria Litovchenko, Elena Smekalova, Alexey Golovizin, Alex Zhavoronkov
2014 Computational and Mathematical Methods in Medicine  
Multilabel classification is often hindered by incompletely labeled training datasets; for some items of such dataset (or even for all of them) some labels may be omitted.  ...  We tried SVM and RF classifiers for the original datasets and then for the modified ones.  ...  They would like to thank the reviewers for many constructive and meaningful comments and suggestions that helped improve the paper and laid the foundation for further research.  ... 
doi:10.1155/2014/781807 pmid:24587817 pmcid:PMC3920912 fatcat:xj5w5rojdzgjpp3iflri6vh2ba

Pointed Subspace Approach to Incomplete Data

Lukasz Struski, Marek Śmieja, Jacek Tabor
2019 Journal of Classification  
To use our representation in practical classification tasks, we embed such generalized missing data into a vector space and define the scalar product of embedding space.  ...  In this paper, we generalize this approach and represent incomplete data as pointed affine subspaces.  ...  Scalar Product for SVM Kernel To apply most of classification methods, it is necessary to define a scalar product (kernel matrix) on a data space.  ... 
doi:10.1007/s00357-019-9304-3 fatcat:nqxtzqg36vckjpplilbmj5nzsy

Sparse Representation Based SAR Vehicle Recognition along with Aspect Angle

Xiangwei Xing, Kefeng Ji, Huanxin Zou, Jixiang Sun
2014 The Scientific World Journal  
As a method of representing the test sample with few training samples from an overcomplete dictionary, sparse representation classification (SRC) has attracted much attention in synthetic aperture radar  ...  In this paper, we develop a novel SAR vehicle recognition method based on sparse representation classification along with aspect information (SRCA), in which the correlation between the vehicle's aspect  ...  For comparison, we compare with several state-of-the-art classification methods: linear SVM, kernel SVM (KSVM) with radial basis function (RBF) kernel, and SRC.  ... 
doi:10.1155/2014/834140 pmid:25161398 pmcid:PMC4000665 fatcat:qda3bbtysjdunihibnqpk4hywy

Why Does Synthesized Data Improve Multi-sequence Classification? [chapter]

Gijs van Tulder, Marleen de Bruijne
2015 Lecture Notes in Computer Science  
Training on the hidden representation from the RBM brought the accuracy of the linear SVMs close to that of random forests.  ...  We present experiments with two classifiers, linear support vector machines (SVMs) and random forests, together with two synthesis methods that can replace missing data in an image classification problem  ...  This research is financed by the Netherlands Organization for Scientific Research (NWO).  ... 
doi:10.1007/978-3-319-24553-9_65 fatcat:3duysw43hbhtvnkv3xokx3ssmm

Upgrading Pulse Detection with Time Shift Properties Using Wavelets and Support Vector Machines [article]

Jaime Gomez, Ignacio Melgar, Juan Seijas
2005 arXiv   pre-print
Each window can contain the pulsed signal (either complete or incomplete) and / or noise.  ...  For each one, the corresponding above-defined SVM was executed. In figure 2 we described how one processing window contains a subset of sampled input data.  ...  For each generated pulse observation, we computed 3 similar Wavelet + Decisionfunction schemes applied to the complete pulse, incomplete 11-samples-shift pulse, and incomplete 23-samples-shift pulse respectively  ... 
arXiv:cs/0505052v1 fatcat:vnt63mudpnfgbce6uzsxtn3gpy

Multi-Hypergraph Learning for Incomplete Multimodality Data

Mingxia Liu, Yue Gao, Pew-Thian Yap, Dinggang Shen
2018 IEEE journal of biomedical and health informatics  
To address these issues, in this paper, we propose a multihypergraph learning (MHL) method for dealing with incomplete multi-modality data.  ...  However, effectively integrating multi-modality data remains a challenging problem especially when the data are incomplete.  ...  Data used in preparation of this article were obtained from the Alzheimer's Disease Neuroimaging Initiative (ADNI) database.  ... 
doi:10.1109/jbhi.2017.2732287 pmid:28749360 pmcid:PMC5785578 fatcat:mfmtftuqvnbs5pmodqhjbip5di

Regression SVM for Incomplete Data

2017 Schedae Informaticae  
This representation allows to construct an analogue of RBF kernel for incomplete data. We show that such a kernel can be successfully used in regression SVM.  ...  The use of machine learning methods in the case of incomplete data is an important task in many scientific fields, like medicine, biology, or face recognition.  ...  Experimental results We applied our probabilistic representation of incomplete data to r-SVM and compared it with various imputation-based techniques.  ... 
doi:10.4467/20838476si.17.001.6807 fatcat:mcayvpe3uff53at6pfsoofctby

Feature Learning from Incomplete EEG with Denoising Autoencoder [article]

Junhua Li, Zbigniew Struzik, Liqing Zhang, Andrzej Cichocki
2014 arXiv   pre-print
The proposed method is evaluated with motor imagery EEG data. The results show that our method can successfully decode incomplete EEG to good effect.  ...  (DAE) for learning.  ...  Therefore, the classification performance using the proposed method for incomplete segments is acceptable for a BCI application 220 system.  ... 
arXiv:1410.0818v1 fatcat:ht63hoqyjzdc7gfsr3nz63x6ly

Region-based automatic web image selection

Keiji Yanai, Kobus Barnard
2010 Proceedings of the international conference on Multimedia information retrieval - MIR '10  
By this region-based classification, we can separate foreground regions from background regions and achieve more effective image training from incomplete training data.  ...  The contribution of this work is (1) to introduce the region-based bag-of-features representation into an Web image selection task where training data is incomplete, and (2) to prove its effectiveness  ...  The novelty of this work is to introduce the region-based bag-of-features representation into an Web image gathering task where training data is incomplete.  ... 
doi:10.1145/1743384.1743436 dblp:conf/mir/YanaiB10 fatcat:4etnxhwyejdcjffprkvmtkmsgy

A Decoding Scheme for Incomplete Motor Imagery EEG With Deep Belief Network

Yaqi Chu, Xingang Zhao, Yijun Zou, Weiliang Xu, Jianda Han, Yiwen Zhao
2018 Frontiers in Neuroscience  
This scheme can provide an alternative decoding solution for the motor imagery EEG contaminated by extreme artifacts and data loss.  ...  Welch and LSP) and two classifiers (DBN and support vector machine, SVM).  ...  The authors would like to thank Huibin Du et al. for participating the experiment.  ... 
doi:10.3389/fnins.2018.00680 pmid:30323737 pmcid:PMC6172343 fatcat:homli5agsrhcvdnpvy7zrzp5ay
« Previous Showing results 1 — 15 out of 10,136 results