Filters








19,905 Hits in 6.5 sec

Fast and accurate text classification via multiple linear discriminant projections

Soumen Chakrabarti, Shourya Roy, Mahesh V. Soundalgekar
2003 The VLDB journal  
Support vector machines (SVMs) have shown superb performance for text classification tasks. They are accurate, robust, and quick to apply to test instances.  ...  It uses Fisher's linear discriminant, a classical tool from statistical pattern recognition, to project training instances to a carefully selected low-dimensional subspace before inducing a decision tree  ...  Acknowledgments: Thanks to Pedro Domingos for helpful discussions, Thorsten Joachims for generous help with SVMlight, Kunal Punera for help with preparing some data sets, and Shantanu Godbole for helpful  ... 
doi:10.1007/s00778-003-0098-9 fatcat:up5r74336zhhziw6wjf3ksctoy

Fast and accurate text classification via multiple linear discriminant projections [chapter]

Soumen Chakrabarti, Shourya Roy, Mahesh V. Soundalgekar
2002 VLDB '02: Proceedings of the 28th International Conference on Very Large Databases  
Support vector machines (SVMs) have shown superb performance for text classification tasks. They are accurate, robust, and quick to apply to test instances.  ...  It uses Fisher's linear discriminant, a classical tool from statistical pattern recognition, to project training instances to a carefully selected low-dimensional subspace before inducing a decision tree  ...  Acknowledgments: Thanks to Pedro Domingos for helpful discussions, Thorsten Joachims for generous help with SVMlight, Kunal Punera for help with preparing some data sets, and Shantanu Godbole for helpful  ... 
doi:10.1016/b978-155860869-6/50064-0 dblp:conf/vldb/ChakrabartiRS02 fatcat:o5toxzstozayzdsuu6hjzgn7pq

Efficient multi-way text categorization via generalized discriminant analysis

Tao Li, Shenghuo Zhu, Mitsunori Ogihara
2003 Proceedings of the twelfth international conference on Information and knowledge management - CIKM '03  
This paper presents a simple and efficient solution to multi-class text categorization. Classification problems are first formulated as optimization via discriminant analysis.  ...  On the other hand, other techniques naturally extensible to handle multi-class classification are generally not as accurate as SVM.  ...  Acknowledgments This work is supported in part by NSF grants EIA-0080124, DUE-9980943, and EIA-0205061, and NIH grant P30-AG18254.  ... 
doi:10.1145/956863.956924 dblp:conf/cikm/LiZO03 fatcat:zmaokhhhnfh3fiky4k56dsgnyq

Text categorization via generalized discriminant analysis

Tao Li, Shenghuo Zhu, Mitsunori Ogihara
2008 Information Processing & Management  
This paper presents a simple and efficient solution to multi-class text categorization. Classification problems are first formulated as optimization via discriminant analysis.  ...  On the other hand, other techniques naturally extensible to handle multi-class classification are generally not as accurate as SVM.  ...  Acknowledgements This work is supported in part by NSF Grants EIA-0080124, DUE-9980943, and EIA-0205061, and NIH Grant P30-AG18254.  ... 
doi:10.1016/j.ipm.2008.03.005 fatcat:ulkdu4etw5aapj6ap6racaepze

Using discriminant analysis for multi-class classification: an experimental investigation

Tao Li, Shenghuo Zhu, Mitsunori Ogihara
2006 Knowledge and Information Systems  
Our experiments suggest that discriminant analysis provides a fast, efficient yet accurate alternative for general multi-class classification problems.  ...  We evaluate the performance of discriminant analysis on a large collection of benchmark datasets and investigate its usage in text categorization.  ...  Fisher's discriminant Using discriminant analysis for multi-class classification 455 analysis was first described for two-class cases [15] , and can be easily extended to multi-class cases via multiple  ... 
doi:10.1007/s10115-006-0013-y fatcat:iiz7b4aqufgu7h3rwj3u6d6xau

Efficient multi-way text categorization via generalized discriminant analysis

Tao Li, Shenghuo Zhu, Mitsunori Ogihara
2003 Proceedings of the twelfth international conference on Information and knowledge management - CIKM '03  
This paper presents a simple and efficient solution to multi-class text categorization. Classification problems are first formulated as optimization via discriminant analysis.  ...  On the other hand, other techniques naturally extensible to handle multi-class classification are generally not as accurate as SVM.  ...  Acknowledgments This work is supported in part by NSF grants EIA-0080124, DUE-9980943, and EIA-0205061, and NIH grant P30-AG18254.  ... 
doi:10.1145/956923.956924 fatcat:tdh5xqc4rrby7np56v24dels74

On-device Structured and Context Partitioned Projection Networks

Sujith Ravi, Zornitsa Kozareva
2019 Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics  
A challenging problem in on-device text classification is to build highly accurate neural models that can fit in small memory footprint and have low latency.  ...  To address this challenge, we propose an on-device neural network SGNN++ which dynamically learns compact projection vectors from raw text using structured and context-dependent partition projections.  ...  Acknowledgments We would like to thank the organizers of the customer feedback challenging for sharing the data and the anonymous reviewers for their valuable feedback and suggestions.  ... 
doi:10.18653/v1/p19-1368 dblp:conf/acl/RaviK19 fatcat:26tnhtzj2nbhtlmkb2k6s64lou

On robust image spam filtering via comprehensive visual modeling

Jialie Shen, Robert H. Deng, Zhiyong Cheng, Liqiang Nie, Shuicheng Yan
2015 Pattern Recognition  
It can facilitate more accurate and robust spam classification process with very limited amount of initial training examples.  ...  In addition, a resampling based learning framework is developed to effectively integrate random forest and linear discriminative analysis (LDA) to generate comprehensive signature of spam images.  ...  The key novelty of LDF is to apply feature selection over subsets of raw features and try to reconstruct a more comprehensive feature combination for superior classification performance via projection.  ... 
doi:10.1016/j.patcog.2015.02.027 fatcat:7bxwrdqm4zggncvpn3dj7ifdwi

A survey of dimensionality reduction techniques based on random projection [article]

Haozhe Xie, Jie Li, Hanqing Xue
2018 arXiv   pre-print
Traditional dimensionality reduction approaches, such as principal component analysis (PCA) and linear discriminant analysis (LDA), have been studied extensively in the past few decades.  ...  These drawbacks have triggered the development of random projection (RP) techniques, which map high-dimensional data onto a low-dimensional subspace with extremely reduced time cost.  ...  Next, K NN classifiers are trained, and the final classification result is produced via a voting scheme. The proposed method is more accurate than the non-ensemble NN classifier. Zhang et al.  ... 
arXiv:1706.04371v4 fatcat:vpdvmo7uffdbhj7cszwj4hud2y

Landmark Classification With Hierarchical Multi-Modal Exemplar Feature

Lei Zhu, Jialie Shen, Hai Jin, Liang Xie, Ran Zheng
2015 IEEE transactions on multimedia  
Then, at the stage of exemplar selection, hierarchical discriminative exemplars in multiple modalities are discovered automatically via iterative boosting and latent region label mining.  ...  The final HMME enjoys advantages of discriminative and linearly separable.  ...  ACKNOWLEDGMENT The authors would like to thank the anonymous reviewers for their constructive and helpful suggestions.  ... 
doi:10.1109/tmm.2015.2431496 fatcat:hhyhhp2ctzbzznm5sfidiiv7ey

Author Index

2010 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition  
Integral Linear Classifier Demo: Fast and Robust Object Segmentation with the Integral Linear Classifier Tome, Pedro Workshop: Scenario-Based Score Fusion for Face Recognition at a Distance Toshev  ...  D Dai , Dai Dengxin Discovering Scene Categories by Information Projection and Cluster Sampling Daniilidis, Kostas Object Detection via Boundary Structure Segmentation Workshop: Video-based Localization  ... 
doi:10.1109/cvpr.2010.5539913 fatcat:y6m5knstrzfyfin6jzusc42p54

Review of Dimension Reduction Methods

Salifu Nanga, Ahmed Tijani Bawah, Benjamin Ansah Acquaye, Mac-Issaka Billa, Francis Delali Baeta, Nii Afotey Odai, Samuel Kwaku Obeng, Ampem Darko Nsiah
2021 Journal of Data Analysis and Information Processing  
Linear Discriminant Analysis (LDA) Linear Discriminant Analysis (LDA) is a well-known and widely used supervised LDRT invented by [140] , who used it successfully for the classification of flowers in  ...  A Multiple Manifold LLE proposed by [252] is an approach that allows for learning multiple manifolds for multiple classes and is efficient in classification and objects recognition.  ... 
doi:10.4236/jdaip.2021.93013 fatcat:tlgvjk6xzbfe5gristkd7ww4tq

Text detection in images using sparse representation with discriminative dictionaries

Ming Zhao, Shutao Li, James Kwok
2010 Image and Vision Computing  
Then, candidate text areas are obtained by applying a simple classification procedure using two learned discriminative dictionaries.  ...  In this paper, we propose a classification-based algorithm for text detection using a sparse representation with discriminative dictionaries.  ...  Acknowledgements The authors would like to thank the editor and anonymous reviewers for their detailed review, valuable comments, and constructive suggestions.  ... 
doi:10.1016/j.imavis.2010.04.002 fatcat:z7ndqrbqpjejvo2ysogcnt6zau

Capturing Long-distance Dependencies in Sequence Models: A Case Study of Chinese Part-of-speech Tagging

Weiwei Sun, Xiaochang Peng, Xiaojun Wan
2013 International Joint Conference on Natural Language Processing  
The re-compiled models not only achieve high accuracy with respect to per token classification, but also serve as a front-end to a parser well.  ...  Second, the structure compilation technique is employed to transfer the predictive power of hybrid models to sequence models via large-scale unlabeled data.  ...  Acknowledgement The work was supported by NSFC (61170166), Beijing Nova Program (2008B03) and National High-Tech R&D Program (2012AA011101).  ... 
dblp:conf/ijcnlp/SunPW13 fatcat:ljdegjsoinc2hatp2y7mr7zjji

Translingual Document Representations from Discriminative Projections

John C. Platt, Kristina Toutanova, Wen-tau Yih
2010 Conference on Empirical Methods in Natural Language Processing  
We use discriminative training to create a projection of documents from multiple languages into a single translingual vector space.  ...  We evaluate these algorithms on two tasks: parallel document retrieval for Wikipedia and Europarl documents, and cross-lingual text classification on Reuters.  ...  When full MT is not practical, a fast word-byword translation model can be used instead, (Ballesteros and Croft, 1996) but may be less accurate.  ... 
dblp:conf/emnlp/PlattTY10 fatcat:fbue2jcsi5gpljxrzvy3ws23ry
« Previous Showing results 1 — 15 out of 19,905 results