combining lsi with other classifiers to improve accuracy
play

Combining LSI with other Classifiers to Improve Accuracy of - PowerPoint PPT Presentation

Jan 02, 2023 •304 likes •745 views

Combining LSI with other Classifiers to Improve Accuracy of Single-label Text Categorization Ana Cardoso-Cachopo Arlindo Oliveira Instituto Superior T ecnico Technical University of Lisbon / INESC-ID EWLSATEL, March 2007

  1. Combining LSI with other Classifiers to Improve Accuracy of Single-label Text Categorization Ana Cardoso-Cachopo Arlindo Oliveira Instituto Superior T´ ecnico — Technical University of Lisbon / INESC-ID EWLSATEL, March 2007 (IST-TULisbon/INESC-ID) Ana Cardoso-Cachopo EWLSATEL, March 2007 1 / 11

  2. Outline Introduction 1 Classification Methods 2 Combinations Between Methods 3 Experimental Setup 4 Experimental Results 5 Conclusions and Future Work 6 (IST-TULisbon/INESC-ID) Ana Cardoso-Cachopo EWLSATEL, March 2007 2 / 11

  3. Outline Introduction 1 Classification Methods 2 Combinations Between Methods 3 Experimental Setup 4 Experimental Results 5 Conclusions and Future Work 6 (IST-TULisbon/INESC-ID) Ana Cardoso-Cachopo EWLSATEL, March 2007 2 / 11

  4. Outline Introduction 1 Classification Methods 2 Combinations Between Methods 3 Experimental Setup 4 Experimental Results 5 Conclusions and Future Work 6 (IST-TULisbon/INESC-ID) Ana Cardoso-Cachopo EWLSATEL, March 2007 2 / 11

  5. Outline Introduction 1 Classification Methods 2 Combinations Between Methods 3 Experimental Setup 4 Experimental Results 5 Conclusions and Future Work 6 (IST-TULisbon/INESC-ID) Ana Cardoso-Cachopo EWLSATEL, March 2007 2 / 11

  6. Outline Introduction 1 Classification Methods 2 Combinations Between Methods 3 Experimental Setup 4 Experimental Results 5 Conclusions and Future Work 6 (IST-TULisbon/INESC-ID) Ana Cardoso-Cachopo EWLSATEL, March 2007 2 / 11

  7. Outline Introduction 1 Classification Methods 2 Combinations Between Methods 3 Experimental Setup 4 Experimental Results 5 Conclusions and Future Work 6 (IST-TULisbon/INESC-ID) Ana Cardoso-Cachopo EWLSATEL, March 2007 2 / 11

  8. Introduction Text Classification Single-label Classification Methods ◮ Vector ◮ k-NN ◮ SVM ◮ LSI Goal: improve Accuracy (IST-TULisbon/INESC-ID) Ana Cardoso-Cachopo EWLSATEL, March 2007 3 / 11

  9. Introduction Text Classification Single-label Classification Methods ◮ Vector ◮ k-NN ◮ SVM ◮ LSI Goal: improve Accuracy (IST-TULisbon/INESC-ID) Ana Cardoso-Cachopo EWLSATEL, March 2007 3 / 11

  10. Introduction Text Classification Single-label Classification Methods ◮ Vector ◮ k-NN ◮ SVM ◮ LSI Goal: improve Accuracy (IST-TULisbon/INESC-ID) Ana Cardoso-Cachopo EWLSATEL, March 2007 3 / 11

  11. Introduction Text Classification Single-label Classification Methods ◮ Vector ◮ k-NN ◮ SVM ◮ LSI Goal: improve Accuracy (IST-TULisbon/INESC-ID) Ana Cardoso-Cachopo EWLSATEL, March 2007 3 / 11

  12. Introduction Text Classification Single-label Classification Methods ◮ Vector ◮ k-NN ◮ SVM ◮ LSI Goal: improve Accuracy (IST-TULisbon/INESC-ID) Ana Cardoso-Cachopo EWLSATEL, March 2007 3 / 11

  13. Introduction Text Classification Single-label Classification Methods ◮ Vector ◮ k-NN ◮ SVM ◮ LSI Goal: improve Accuracy (IST-TULisbon/INESC-ID) Ana Cardoso-Cachopo EWLSATEL, March 2007 3 / 11

  14. Introduction Text Classification Single-label Classification Methods ◮ Vector ◮ k-NN ◮ SVM ◮ LSI Goal: improve Accuracy (IST-TULisbon/INESC-ID) Ana Cardoso-Cachopo EWLSATEL, March 2007 3 / 11

  15. Introduction Text Classification Single-label Classification Methods ◮ Vector ◮ k-NN ◮ SVM ◮ LSI Goal: improve Accuracy (IST-TULisbon/INESC-ID) Ana Cardoso-Cachopo EWLSATEL, March 2007 3 / 11

  16. Classification Methods Cosine similarity Vector k-NN + Cosine similarity k-NN Voting strategy Kernel SVM p dimensional term space Cosine similarity SVD LSI s << p dimensional concept space (IST-TULisbon/INESC-ID) Ana Cardoso-Cachopo EWLSATEL, March 2007 4 / 11

  17. Combinations Between Methods k-NN + Cosine similarity k-NN-LSI SVD Kernel + Voting strategy SVM-LSI s << p p dimensional dimensional term space concept space (IST-TULisbon/INESC-ID) Ana Cardoso-Cachopo EWLSATEL, March 2007 5 / 11

  18. Experimental Setup Methods (6 already mentioned + Dumb) Datasets ◮ Bank´s Data - Bank37 ◮ Reuters 21578 - R8, R52 ◮ 20 Newsgroups - 20Ng ◮ Web Knowledge Base - Web4 ◮ Cade - Cade12 Evaluation Measure Accuracy = #Correctly classified documents #Total documents (IST-TULisbon/INESC-ID) Ana Cardoso-Cachopo EWLSATEL, March 2007 6 / 11

  19. Experimental Setup Methods (6 already mentioned + Dumb) Datasets ◮ Bank´s Data - Bank37 ◮ Reuters 21578 - R8, R52 ◮ 20 Newsgroups - 20Ng ◮ Web Knowledge Base - Web4 ◮ Cade - Cade12 Evaluation Measure Accuracy = #Correctly classified documents #Total documents (IST-TULisbon/INESC-ID) Ana Cardoso-Cachopo EWLSATEL, March 2007 6 / 11

  20. Experimental Setup Methods (6 already mentioned + Dumb) Datasets ◮ Bank´s Data - Bank37 ◮ Reuters 21578 - R8, R52 ◮ 20 Newsgroups - 20Ng ◮ Web Knowledge Base - Web4 ◮ Cade - Cade12 Evaluation Measure Accuracy = #Correctly classified documents #Total documents (IST-TULisbon/INESC-ID) Ana Cardoso-Cachopo EWLSATEL, March 2007 6 / 11

  21. Experimental Setup Methods (6 already mentioned + Dumb) Datasets ◮ Bank´s Data - Bank37 ◮ Reuters 21578 - R8, R52 ◮ 20 Newsgroups - 20Ng ◮ Web Knowledge Base - Web4 ◮ Cade - Cade12 Evaluation Measure Accuracy = #Correctly classified documents #Total documents (IST-TULisbon/INESC-ID) Ana Cardoso-Cachopo EWLSATEL, March 2007 6 / 11

  22. Experimental Setup Methods (6 already mentioned + Dumb) Datasets ◮ Bank´s Data - Bank37 ◮ Reuters 21578 - R8, R52 ◮ 20 Newsgroups - 20Ng ◮ Web Knowledge Base - Web4 ◮ Cade - Cade12 Evaluation Measure Accuracy = #Correctly classified documents #Total documents (IST-TULisbon/INESC-ID) Ana Cardoso-Cachopo EWLSATEL, March 2007 6 / 11

  23. Experimental Setup Methods (6 already mentioned + Dumb) Datasets ◮ Bank´s Data - Bank37 ◮ Reuters 21578 - R8, R52 ◮ 20 Newsgroups - 20Ng ◮ Web Knowledge Base - Web4 ◮ Cade - Cade12 Evaluation Measure Accuracy = #Correctly classified documents #Total documents (IST-TULisbon/INESC-ID) Ana Cardoso-Cachopo EWLSATEL, March 2007 6 / 11

  24. Experimental Setup Methods (6 already mentioned + Dumb) Datasets ◮ Bank´s Data - Bank37 ◮ Reuters 21578 - R8, R52 ◮ 20 Newsgroups - 20Ng ◮ Web Knowledge Base - Web4 ◮ Cade - Cade12 Evaluation Measure Accuracy = #Correctly classified documents #Total documents (IST-TULisbon/INESC-ID) Ana Cardoso-Cachopo EWLSATEL, March 2007 6 / 11

  25. Experimental Setup Methods (6 already mentioned + Dumb) Datasets ◮ Bank´s Data - Bank37 ◮ Reuters 21578 - R8, R52 ◮ 20 Newsgroups - 20Ng ◮ Web Knowledge Base - Web4 ◮ Cade - Cade12 Evaluation Measure Accuracy = #Correctly classified documents #Total documents (IST-TULisbon/INESC-ID) Ana Cardoso-Cachopo EWLSATEL, March 2007 6 / 11

  26. Characteristics of the Datasets Train Test Total Smallest Largest Docs Docs Docs Class Class Bank37 928 463 1391 5 346 20Ng 11293 7528 18821 628 999 R8 5485 2189 7674 51 3923 R52 6532 2568 9100 3 3923 Web4 2803 1396 4199 504 1641 Cade12 27322 13661 40983 625 8473 Numbers of documents for the datasets: number of training documents, number of test documents, total number of documents, number of documents in the smallest class, and number of documents in the largest class. (IST-TULisbon/INESC-ID) Ana Cardoso-Cachopo EWLSATEL, March 2007 7 / 11

  27. Characteristics of the Datasets Train Test Total Smallest Largest Docs Docs Docs Class Class Bank37 928 463 1391 5 346 20Ng 11293 7528 18821 628 999 R8 5485 2189 7674 51 3923 R52 6532 2568 9100 3 3923 Web4 2803 1396 4199 504 1641 Cade12 27322 13661 40983 625 8473 Numbers of documents for the datasets: number of training documents, number of test documents, total number of documents, number of documents in the smallest class, and number of documents in the largest class. (IST-TULisbon/INESC-ID) Ana Cardoso-Cachopo EWLSATEL, March 2007 7 / 11

  28. Characteristics of the Datasets Train Test Total Smallest Largest Docs Docs Docs Class Class Bank37 928 463 1391 5 346 20Ng 11293 7528 18821 628 999 R8 5485 2189 7674 51 3923 R52 6532 2568 9100 3 3923 Web4 2803 1396 4199 504 1641 Cade12 27322 13661 40983 625 8473 Numbers of documents for the datasets: number of training documents, number of test documents, total number of documents, number of documents in the smallest class, and number of documents in the largest class. (IST-TULisbon/INESC-ID) Ana Cardoso-Cachopo EWLSATEL, March 2007 7 / 11

  29. Characteristics of the Datasets Train Test Total Smallest Largest Docs Docs Docs Class Class Bank37 928 463 1391 5 346 20Ng 11293 7528 18821 628 999 R8 5485 2189 7674 51 3923 R52 6532 2568 9100 3 3923 Web4 2803 1396 4199 504 1641 Cade12 27322 13661 40983 625 8473 Numbers of documents for the datasets: number of training documents, number of test documents, total number of documents, number of documents in the smallest class, and number of documents in the largest class. (IST-TULisbon/INESC-ID) Ana Cardoso-Cachopo EWLSATEL, March 2007 7 / 11

  30. Experimental Results 1.0 Dumb Vector 0.8 k-NN SVM 0.6 LSI k-NN-LSI 0.4 SVM-LSI 0.2 0.0 Bank37 20Ng R8 R52 Web4 Cade12 Accuracy values for the six datasets using each method. (IST-TULisbon/INESC-ID) Ana Cardoso-Cachopo EWLSATEL, March 2007 8 / 11

Recommend

Combining Classifiers d i,j = 1 if D i labels x in i , and d i,j = 0 otherwise. In this case,

Combining Classifiers d i,j = 1 if D i labels x in i , and d i,j = 0 otherwise. In this case,

Majority Vote Classifiers output a c-dimensional binary vector [ d i, 1 , ..., d i,c ] T { 0 , 1 } c , where i = 1 , ..., L and Combining Classifiers d i,j = 1 if D i labels x in i , and d i,j = 0 otherwise. In this case, plurality will

630 views • 4 slides
Data Dependence in Data Dependence in Combining Classifiers Combining Classifiers Mohamed

Data Dependence in Data Dependence in Combining Classifiers Combining Classifiers Mohamed

Data Dependence in Data Dependence in Combining Classifiers Combining Classifiers Mohamed Kamel, Nayer Wanas Mohamed Kamel, Nayer Wanas Pattern Analysis and Machine Intelligence Lab Pattern Analysis and Machine Intelligence Lab University of

690 views • 38 slides
Combining Heterogeneous Classifiers for Word-Sense Disambiguation Dan Klein , Kristina Toutanova ,

Combining Heterogeneous Classifiers for Word-Sense Disambiguation Dan Klein , Kristina Toutanova ,

Proceedings of the SIGLEX/SENSEVAL Workshop on Word Sense Disambiguation: Recent Successes and Future Directions, Philadelphia, July 2002, pp. 74-80. Association for Computational Linguistics. Combining Heterogeneous Classifiers for Word-Sense

381 views • 7 slides
Bounding the fairness and accuracy of classifiers from population statistics ICML 2020 Sivan

Bounding the fairness and accuracy of classifiers from population statistics ICML 2020 Sivan

Bounding the fairness and accuracy of classifiers from population statistics ICML 2020 Sivan Sabato and Elad Yom-Tov Sabato &amp; Yom-Tov (Microsoft &amp; BGU) Bounding fairness and accuracy 1 / 15 The 1-slide summary We show how to study a

661 views • 52 slides
Combining Classifiers: A Theoretical Framework J. Kittler Centre for Vision, Speech and Signal

Combining Classifiers: A Theoretical Framework J. Kittler Centre for Vision, Speech and Signal

Pattern Analysis &amp; Applic. (1998)1:18-27 9 1998 Springer-Verlag London Limited Combining Classifiers: A Theoretical Framework J. Kittler Centre for Vision, Speech and Signal Processing, School of Electronic Engineering, Information

464 views • 10 slides
Training to Improve Judgmental Expertise by Using Decompositions of Judgment Accuracy Measures

Training to Improve Judgmental Expertise by Using Decompositions of Judgment Accuracy Measures

Training to Improve Judgmental Expertise by Using Decompositions of Judgment Accuracy Measures Eric R. Stone Wake Forest University Overarching Goal How can we improve judgment accuracy? Two types of judgments: 1. Judgments of discrete

1.08k views • 59 slides
Nonlinear Classifiers II 2 Nonlinear Classifiers: Introduction Classifiers Supervised

Nonlinear Classifiers II 2 Nonlinear Classifiers: Introduction Classifiers Supervised

1 Nonlinear Classifiers II 2 Nonlinear Classifiers: Introduction Classifiers Supervised Classifiers XOR problem Linear Classifiers Perceptron Least Squares Methods Linear Support Vector Machine Nonlinear

355 views • 20 slides
Classifiers that improve with use Argument George Nagy In-house training sets are never DocLab

Classifiers that improve with use Argument George Nagy In-house training sets are never DocLab

PRMU Classifiers that improve with use Argument George Nagy In-house training sets are never DocLab large enough, and never representative enough. Rensselaer Polytechnic Institute We must therefore augment them with samples from actual

570 views • 11 slides
COS424 Scribe Notes Lecture 14: Ensembles Donghun Lee April 8, 2010 1 Ensembles A set of

COS424 Scribe Notes Lecture 14: Ensembles Donghun Lee April 8, 2010 1 Ensembles A set of

COS424 Scribe Notes Lecture 14: Ensembles Donghun Lee April 8, 2010 1 Ensembles A set of classifiers can combine their outputs to make an ensemble of the classifiers. Two require- ments of such an ensemble to work: accuracy: classifiers

286 views • 4 slides
Optimally Combining Outcomes to Improve Prediction David Benkeser benkeser@berkeley.edu UC

Optimally Combining Outcomes to Improve Prediction David Benkeser benkeser@berkeley.edu UC

Optimally Combining Outcomes to Improve Prediction David Benkeser benkeser@berkeley.edu UC Berkeley Biostatistics November 15, 2016 Acknowledgments Collaborators: Mark van der Laan, Alan Hubbard, Ben Arnold, Jack Colford, Andrew Mertens,

863 views • 60 slides
Nonlinear Dimension Reduction to Improve Predictive Accuracy in Genomic and Neuroimaging Studies

Nonlinear Dimension Reduction to Improve Predictive Accuracy in Genomic and Neuroimaging Studies

Nonlinear Dimension Reduction to Improve Predictive Accuracy in Genomic and Neuroimaging Studies Maxime Turgeon June 5, 2018 McGill University Department of Epidemiology, Biostatistics, and Occupational Health 1/21 Acknowledgements This

514 views • 25 slides
Combining GOES R and GPM to improve GOES R rainrate product Nai Yu Wang, University of

Combining GOES R and GPM to improve GOES R rainrate product Nai Yu Wang, University of

Combining GOES R and GPM to improve GOES R rainrate product Nai Yu Wang, University of Maryland, ESSIC/CICS Kaushik Gopalan, ISRO, India Rachel Albrecht, CPTEC, Brazil Weixin Xu, University of Maryland, ESSIC/CICS Rober Adler, University

566 views • 19 slides
Effect of Classifiers in Consensus Feature Ranking for Biomedical Datasets Dimension Reduction

Effect of Classifiers in Consensus Feature Ranking for Biomedical Datasets Dimension Reduction

Shobeir Fakhraei, Hamid Soltanian-Zadeh, Farshad Fotouhi, Kost Elisevich Effect of Classifiers in Consensus Feature Ranking for Biomedical Datasets Dimension Reduction Prediction accuracy of practical machine learning algorithms degrades

510 views • 22 slides
Ants, Mutants and Beyond Combining formal and stochastic techniques to improve software

Ants, Mutants and Beyond Combining formal and stochastic techniques to improve software

Nature-inspired Computing Search Based Software Engineering Genetic Improvement Automatic Improvement References References Ants, Mutants and Beyond Combining formal and stochastic techniques to improve software jerry.swan@york.ac.uk

487 views • 26 slides
Occasion-level Classifiers or Event-level Classifiers? -Evidence from Child Language Acquisition

Occasion-level Classifiers or Event-level Classifiers? -Evidence from Child Language Acquisition

Occasion-level Classifiers or Event-level Classifiers? -Evidence from Child Language Acquisition Yuan Xiaohe, Zhang Xiaoqian, Luo Yingyi, Hu Jianhua Institute of Linguistics Chinese Academy of Social Sciences Verbal classifiers in Mandarin

823 views • 24 slides
Machine Learning Nave Bayes classifiers Types of classifiers We can divide the large

Machine Learning Nave Bayes classifiers Types of classifiers We can divide the large

10-701 Machine Learning Nave Bayes classifiers Types of classifiers We can divide the large variety of classification approaches into three major types 1. Instance based classifiers - Use observation directly (no models) - e.g. K

309 views • 28 slides
When a Tree Falls: Using Diversity in Ensemble Classifiers to Identify Evasion in Malware

When a Tree Falls: Using Diversity in Ensemble Classifiers to Identify Evasion in Malware

When a Tree Falls: Using Diversity in Ensemble Classifiers to Identify Evasion in Malware Detectors Charles Smutz Angelos Stavrou George Mason University Motivation Machine learning used ubiquitously to improve information security

814 views • 30 slides
Minimal Cost Complexity Pruning of Meta-Classifiers Andreas L. Prodromidis Salvatore J. Stolfo

Minimal Cost Complexity Pruning of Meta-Classifiers Andreas L. Prodromidis Salvatore J. Stolfo

Minimal Cost Complexity Pruning of Meta-Classifiers Andreas L. Prodromidis Salvatore J. Stolfo Department of Computer Science Columbia University Combining multiple models Learning Algorithm Classifier-1 Learning Training Classifier-2

372 views • 18 slides
Monitor and Improve the GOES Imager Infrared Calibration Accuracy: An Application of GSICS

Monitor and Improve the GOES Imager Infrared Calibration Accuracy: An Application of GSICS

Monitor and Improve the GOES Imager Infrared Calibration Accuracy: An Application of GSICS Fangfang Yu 1 , Xiangqian Fred Wu 2 and Mitch Goldberg 2 1: ERT, Inc @ NOAA/NESDIS/STAR 2: NOAA/NESDIS/STAR 11 November 2011 GSICS GEO-LEO

412 views • 18 slides
Combining Solr and Elasticsearch to Improve Autosuggestion on Mobile Local Search Toan Vinh Luu,

Combining Solr and Elasticsearch to Improve Autosuggestion on Mobile Local Search Toan Vinh Luu,

Combining Solr and Elasticsearch to Improve Autosuggestion on Mobile Local Search Toan Vinh Luu, PhD Senior Search Engineer local.ch AG Apache: Big Data 2015 In this talk Requirements of an autosuggestion feature Autosuggestion

416 views • 30 slides
NYU WIRELESS The World's First Academic Research Center Combining Wireless, Computing, and Medical

NYU WIRELESS The World's First Academic Research Center Combining Wireless, Computing, and Medical

NYU WIRELESS The World's First Academic Research Center Combining Wireless, Computing, and Medical Applications Investigation of Prediction Accuracy, Sensitivity, and Parameter Stability of Large-Scale Propagation Path Loss Models for 5G

222 views • 10 slides
Automatically Evading Classifiers A Case Study on PDF Malware Classifiers Weilin Xu

Automatically Evading Classifiers A Case Study on PDF Malware Classifiers Weilin Xu

Automatically Evading Classifiers A Case Study on PDF Malware Classifiers Weilin Xu David Evans Yanjun Qi University of Virginia Machine Learning is Solving Our Problems Fake Fake Spam IDS

1.07k views • 34 slides
CS440/ECE448 Lecture 22: Including Slides by Svetlana Lazebnik, 10/2016 Linear Classifiers

CS440/ECE448 Lecture 22: Including Slides by Svetlana Lazebnik, 10/2016 Linear Classifiers

Mark Hasegawa-Johnson, 3/2020 CS440/ECE448 Lecture 22: Including Slides by Svetlana Lazebnik, 10/2016 Linear Classifiers License: CC-BY 4.0 Linear Classifiers Classifiers Perceptron Linear classifiers in general Logistic

581 views • 45 slides
Cognitive Modeling Unseen Examples 2 Bayes Classifiers Lecture 14: Naive Bayes Classifiers

Cognitive Modeling Unseen Examples 2 Bayes Classifiers Lecture 14: Naive Bayes Classifiers

Decision Making as Classification Decision Making as Classification Bayes Classifiers Bayes Classifiers Naive Bayes Classifiers Naive Bayes Classifiers 1 Decision Making as Classification Decision Making Frequencies and Probabilities

268 views • 5 slides

More recommend