SlideShare a Scribd company logo
1 of 13
Self-taught learning: Transfer learning from unlabelled data Authors: Rajat Raina et al. @Stanford University Presenter: Shao-Chuan Wang Some slides are from author’s work
Self-taught learning: Transfer learning from unlabelled data Objective: Use the unlabelled data to improve the performance on a classification task. Kea Ideas:  Relax the assumption about the unlabelled data. Use unlabelled data to learn the best representation (dictionary)
Machine Learning Schemes ,[object Object]
Semi-supervised learning.
Transfer learning.
Next: Self-taught learning?Cars Motorcycles Cars Motorcycles Cars Motorcycles Bus Tractor Aircraft Helicopter Motorcycle Car Natural scenes
Self-taught Learning Labeled examples:  Unlabeled examples: The unlabeled and labeled data: Need not share labels y. Need not share a generative distribution. Advantage: Such unlabeled data is often easy to obtain.
Self-taught learning Use unlabelled data to learn the best representation.
Learning the structure Sparse coding: learning the dictionary Encoding (L1-regularized least square problem) Least angle regression (Efron et al. 2004) (very fast!!) Feature-sign search (Honglak Lee et al. 2006) Coordinate descent (Friedman et al. 2008) Update dictionary(L2-constrained least square problem) K-SVD (Aharon et al. 2006) Online dictionary learning (Mairal et al. 2009) Reconstruction error Sparsity penalty
Self-taught learning: flow K-SVD, online dictionary learning
 Learning algorithm Unlabelled data Dictionary Encoded Training features Encoding SVM with Fisher kernel Training data Encoded Testing features Testing data The Lasso Predicted Labels
SVM with Fisher kernel Fisher kernel In Bayesian view, Laplace prior
Example atoms Natural images. “edges” “strokes” Handwritten characters. Sparse representation gives a higher level representation
Result: Character recognition Digits Handwritten English English font Handwritten English classification (20 labeled images per handwritten character) Bases learnt on digits English font classification (20 labeled images per font character) Bases learnt on handwritten English 8.2% error reduction 2.8% error reduction

More Related Content

Similar to Self Taught Learning

Final Presentation V3
Final Presentation V3Final Presentation V3
Final Presentation V3weichen
 
Query Linguistic Intent Detection
Query Linguistic Intent DetectionQuery Linguistic Intent Detection
Query Linguistic Intent Detectionbutest
 
Adaptive membership functions for hand written character recognition by voron...
Adaptive membership functions for hand written character recognition by voron...Adaptive membership functions for hand written character recognition by voron...
Adaptive membership functions for hand written character recognition by voron...JPINFOTECH JAYAPRAKASH
 
Weakly Supervised Machine Reading
Weakly Supervised Machine ReadingWeakly Supervised Machine Reading
Weakly Supervised Machine ReadingIsabelle Augenstein
 
AP Computer Science Test Prep - Part 2 - Object Oriented Programming
AP Computer Science Test Prep - Part 2 - Object Oriented ProgrammingAP Computer Science Test Prep - Part 2 - Object Oriented Programming
AP Computer Science Test Prep - Part 2 - Object Oriented ProgrammingNR Computer Learning Center
 
èŻŸć ‚èźČäč‰(æœ€ćŽæ›Žæ–°:2009-9-25)
èŻŸć ‚èźČäč‰(æœ€ćŽæ›Žæ–°:2009-9-25)èŻŸć ‚èźČäč‰(æœ€ćŽæ›Žæ–°:2009-9-25)
èŻŸć ‚èźČäč‰(æœ€ćŽæ›Žæ–°:2009-9-25)butest
 
NL based Object Oriented modeling - EJSR 35(1)
NL based Object Oriented modeling - EJSR 35(1)NL based Object Oriented modeling - EJSR 35(1)
NL based Object Oriented modeling - EJSR 35(1)IT Industry
 
Introduction to Python
Introduction to PythonIntroduction to Python
Introduction to Pythontswr
 
CVPR2009: Object Detection Using a Max-Margin Hough Transform
CVPR2009: Object Detection Using a Max-Margin Hough TransformCVPR2009: Object Detection Using a Max-Margin Hough Transform
CVPR2009: Object Detection Using a Max-Margin Hough Transformzukun
 

Similar to Self Taught Learning (11)

Final Presentation V3
Final Presentation V3Final Presentation V3
Final Presentation V3
 
Query Linguistic Intent Detection
Query Linguistic Intent DetectionQuery Linguistic Intent Detection
Query Linguistic Intent Detection
 
Adaptive membership functions for hand written character recognition by voron...
Adaptive membership functions for hand written character recognition by voron...Adaptive membership functions for hand written character recognition by voron...
Adaptive membership functions for hand written character recognition by voron...
 
Ontology
OntologyOntology
Ontology
 
Weakly Supervised Machine Reading
Weakly Supervised Machine ReadingWeakly Supervised Machine Reading
Weakly Supervised Machine Reading
 
AP Computer Science Test Prep - Part 2 - Object Oriented Programming
AP Computer Science Test Prep - Part 2 - Object Oriented ProgrammingAP Computer Science Test Prep - Part 2 - Object Oriented Programming
AP Computer Science Test Prep - Part 2 - Object Oriented Programming
 
èŻŸć ‚èźČäč‰(æœ€ćŽæ›Žæ–°:2009-9-25)
èŻŸć ‚èźČäč‰(æœ€ćŽæ›Žæ–°:2009-9-25)èŻŸć ‚èźČäč‰(æœ€ćŽæ›Žæ–°:2009-9-25)
èŻŸć ‚èźČäč‰(æœ€ćŽæ›Žæ–°:2009-9-25)
 
NL based Object Oriented modeling - EJSR 35(1)
NL based Object Oriented modeling - EJSR 35(1)NL based Object Oriented modeling - EJSR 35(1)
NL based Object Oriented modeling - EJSR 35(1)
 
Introduction to Python
Introduction to PythonIntroduction to Python
Introduction to Python
 
sahuPPT.pptx
sahuPPT.pptxsahuPPT.pptx
sahuPPT.pptx
 
CVPR2009: Object Detection Using a Max-Margin Hough Transform
CVPR2009: Object Detection Using a Max-Margin Hough TransformCVPR2009: Object Detection Using a Max-Margin Hough Transform
CVPR2009: Object Detection Using a Max-Margin Hough Transform
 

More from Shao-Chuan Wang

Introduction to Machine Learning
Introduction to Machine LearningIntroduction to Machine Learning
Introduction to Machine LearningShao-Chuan Wang
 
Beyond The Euclidean Distance: Creating effective visual codebooks using the ...
Beyond The Euclidean Distance: Creating effective visual codebooks using the ...Beyond The Euclidean Distance: Creating effective visual codebooks using the ...
Beyond The Euclidean Distance: Creating effective visual codebooks using the ...Shao-Chuan Wang
 
A Friendly Guide To Sparse Coding
A Friendly Guide To Sparse CodingA Friendly Guide To Sparse Coding
A Friendly Guide To Sparse CodingShao-Chuan Wang
 
An Exemplar Model For Learning Object Classes
An Exemplar Model For Learning Object ClassesAn Exemplar Model For Learning Object Classes
An Exemplar Model For Learning Object ClassesShao-Chuan Wang
 
Evaluation Of Color Descriptors For Object And Scene
Evaluation Of Color Descriptors For Object And SceneEvaluation Of Color Descriptors For Object And Scene
Evaluation Of Color Descriptors For Object And SceneShao-Chuan Wang
 
Spatially Coherent Latent Topic Model For Concurrent Object Segmentation and ...
Spatially Coherent Latent Topic Model For Concurrent Object Segmentation and ...Spatially Coherent Latent Topic Model For Concurrent Object Segmentation and ...
Spatially Coherent Latent Topic Model For Concurrent Object Segmentation and ...Shao-Chuan Wang
 
Support Vector Machine
Support Vector MachineSupport Vector Machine
Support Vector MachineShao-Chuan Wang
 
Image Classification And Support Vector Machine
Image Classification And Support Vector MachineImage Classification And Support Vector Machine
Image Classification And Support Vector MachineShao-Chuan Wang
 

More from Shao-Chuan Wang (9)

Introduction to Machine Learning
Introduction to Machine LearningIntroduction to Machine Learning
Introduction to Machine Learning
 
Beyond The Euclidean Distance: Creating effective visual codebooks using the ...
Beyond The Euclidean Distance: Creating effective visual codebooks using the ...Beyond The Euclidean Distance: Creating effective visual codebooks using the ...
Beyond The Euclidean Distance: Creating effective visual codebooks using the ...
 
A Friendly Guide To Sparse Coding
A Friendly Guide To Sparse CodingA Friendly Guide To Sparse Coding
A Friendly Guide To Sparse Coding
 
An Exemplar Model For Learning Object Classes
An Exemplar Model For Learning Object ClassesAn Exemplar Model For Learning Object Classes
An Exemplar Model For Learning Object Classes
 
Evaluation Of Color Descriptors For Object And Scene
Evaluation Of Color Descriptors For Object And SceneEvaluation Of Color Descriptors For Object And Scene
Evaluation Of Color Descriptors For Object And Scene
 
Spatially Coherent Latent Topic Model For Concurrent Object Segmentation and ...
Spatially Coherent Latent Topic Model For Concurrent Object Segmentation and ...Spatially Coherent Latent Topic Model For Concurrent Object Segmentation and ...
Spatially Coherent Latent Topic Model For Concurrent Object Segmentation and ...
 
Support Vector Machine
Support Vector MachineSupport Vector Machine
Support Vector Machine
 
About Python
About PythonAbout Python
About Python
 
Image Classification And Support Vector Machine
Image Classification And Support Vector MachineImage Classification And Support Vector Machine
Image Classification And Support Vector Machine
 

Recently uploaded

PSYCHIATRIC History collection FORMAT.pptx
PSYCHIATRIC   History collection FORMAT.pptxPSYCHIATRIC   History collection FORMAT.pptx
PSYCHIATRIC History collection FORMAT.pptxPoojaSen20
 
Introduction to ArtificiaI Intelligence in Higher Education
Introduction to ArtificiaI Intelligence in Higher EducationIntroduction to ArtificiaI Intelligence in Higher Education
Introduction to ArtificiaI Intelligence in Higher Educationpboyjonauth
 
Introduction to AI in Higher Education_draft.pptx
Introduction to AI in Higher Education_draft.pptxIntroduction to AI in Higher Education_draft.pptx
Introduction to AI in Higher Education_draft.pptxpboyjonauth
 
The Most Excellent Way | 1 Corinthians 13
The Most Excellent Way | 1 Corinthians 13The Most Excellent Way | 1 Corinthians 13
The Most Excellent Way | 1 Corinthians 13Steve Thomason
 
microwave assisted reaction. General introduction
microwave assisted reaction. General introductionmicrowave assisted reaction. General introduction
microwave assisted reaction. General introductionMaksud Ahmed
 
APM Welcome, APM North West Network Conference, Synergies Across Sectors
APM Welcome, APM North West Network Conference, Synergies Across SectorsAPM Welcome, APM North West Network Conference, Synergies Across Sectors
APM Welcome, APM North West Network Conference, Synergies Across SectorsAssociation for Project Management
 
How to Make a Pirate ship Primary Education.pptx
How to Make a Pirate ship Primary Education.pptxHow to Make a Pirate ship Primary Education.pptx
How to Make a Pirate ship Primary Education.pptxmanuelaromero2013
 
Employee wellbeing at the workplace.pptx
Employee wellbeing at the workplace.pptxEmployee wellbeing at the workplace.pptx
Employee wellbeing at the workplace.pptxNirmalaLoungPoorunde1
 
Organic Name Reactions for the students and aspirants of Chemistry12th.pptx
Organic Name Reactions  for the students and aspirants of Chemistry12th.pptxOrganic Name Reactions  for the students and aspirants of Chemistry12th.pptx
Organic Name Reactions for the students and aspirants of Chemistry12th.pptxVS Mahajan Coaching Centre
 
Hybridoma Technology ( Production , Purification , and Application )
Hybridoma Technology  ( Production , Purification , and Application  ) Hybridoma Technology  ( Production , Purification , and Application  )
Hybridoma Technology ( Production , Purification , and Application ) Sakshi Ghasle
 
Sanyam Choudhary Chemistry practical.pdf
Sanyam Choudhary Chemistry practical.pdfSanyam Choudhary Chemistry practical.pdf
Sanyam Choudhary Chemistry practical.pdfsanyamsingh5019
 
Grant Readiness 101 TechSoup and Remy Consulting
Grant Readiness 101 TechSoup and Remy ConsultingGrant Readiness 101 TechSoup and Remy Consulting
Grant Readiness 101 TechSoup and Remy ConsultingTechSoup
 
Mastering the Unannounced Regulatory Inspection
Mastering the Unannounced Regulatory InspectionMastering the Unannounced Regulatory Inspection
Mastering the Unannounced Regulatory InspectionSafetyChain Software
 
Arihant handbook biology for class 11 .pdf
Arihant handbook biology for class 11 .pdfArihant handbook biology for class 11 .pdf
Arihant handbook biology for class 11 .pdfchloefrazer622
 
MENTAL STATUS EXAMINATION format.docx
MENTAL     STATUS EXAMINATION format.docxMENTAL     STATUS EXAMINATION format.docx
MENTAL STATUS EXAMINATION format.docxPoojaSen20
 
Presentation by Andreas Schleicher Tackling the School Absenteeism Crisis 30 ...
Presentation by Andreas Schleicher Tackling the School Absenteeism Crisis 30 ...Presentation by Andreas Schleicher Tackling the School Absenteeism Crisis 30 ...
Presentation by Andreas Schleicher Tackling the School Absenteeism Crisis 30 ...EduSkills OECD
 
SOCIAL AND HISTORICAL CONTEXT - LFTVD.pptx
SOCIAL AND HISTORICAL CONTEXT - LFTVD.pptxSOCIAL AND HISTORICAL CONTEXT - LFTVD.pptx
SOCIAL AND HISTORICAL CONTEXT - LFTVD.pptxiammrhaywood
 
BASLIQ CURRENT LOOKBOOK LOOKBOOK(1) (1).pdf
BASLIQ CURRENT LOOKBOOK  LOOKBOOK(1) (1).pdfBASLIQ CURRENT LOOKBOOK  LOOKBOOK(1) (1).pdf
BASLIQ CURRENT LOOKBOOK LOOKBOOK(1) (1).pdfSoniaTolstoy
 
Paris 2024 Olympic Geographies - an activity
Paris 2024 Olympic Geographies - an activityParis 2024 Olympic Geographies - an activity
Paris 2024 Olympic Geographies - an activityGeoBlogs
 

Recently uploaded (20)

PSYCHIATRIC History collection FORMAT.pptx
PSYCHIATRIC   History collection FORMAT.pptxPSYCHIATRIC   History collection FORMAT.pptx
PSYCHIATRIC History collection FORMAT.pptx
 
Introduction to ArtificiaI Intelligence in Higher Education
Introduction to ArtificiaI Intelligence in Higher EducationIntroduction to ArtificiaI Intelligence in Higher Education
Introduction to ArtificiaI Intelligence in Higher Education
 
Introduction to AI in Higher Education_draft.pptx
Introduction to AI in Higher Education_draft.pptxIntroduction to AI in Higher Education_draft.pptx
Introduction to AI in Higher Education_draft.pptx
 
The Most Excellent Way | 1 Corinthians 13
The Most Excellent Way | 1 Corinthians 13The Most Excellent Way | 1 Corinthians 13
The Most Excellent Way | 1 Corinthians 13
 
microwave assisted reaction. General introduction
microwave assisted reaction. General introductionmicrowave assisted reaction. General introduction
microwave assisted reaction. General introduction
 
APM Welcome, APM North West Network Conference, Synergies Across Sectors
APM Welcome, APM North West Network Conference, Synergies Across SectorsAPM Welcome, APM North West Network Conference, Synergies Across Sectors
APM Welcome, APM North West Network Conference, Synergies Across Sectors
 
How to Make a Pirate ship Primary Education.pptx
How to Make a Pirate ship Primary Education.pptxHow to Make a Pirate ship Primary Education.pptx
How to Make a Pirate ship Primary Education.pptx
 
Employee wellbeing at the workplace.pptx
Employee wellbeing at the workplace.pptxEmployee wellbeing at the workplace.pptx
Employee wellbeing at the workplace.pptx
 
Organic Name Reactions for the students and aspirants of Chemistry12th.pptx
Organic Name Reactions  for the students and aspirants of Chemistry12th.pptxOrganic Name Reactions  for the students and aspirants of Chemistry12th.pptx
Organic Name Reactions for the students and aspirants of Chemistry12th.pptx
 
Hybridoma Technology ( Production , Purification , and Application )
Hybridoma Technology  ( Production , Purification , and Application  ) Hybridoma Technology  ( Production , Purification , and Application  )
Hybridoma Technology ( Production , Purification , and Application )
 
Sanyam Choudhary Chemistry practical.pdf
Sanyam Choudhary Chemistry practical.pdfSanyam Choudhary Chemistry practical.pdf
Sanyam Choudhary Chemistry practical.pdf
 
Grant Readiness 101 TechSoup and Remy Consulting
Grant Readiness 101 TechSoup and Remy ConsultingGrant Readiness 101 TechSoup and Remy Consulting
Grant Readiness 101 TechSoup and Remy Consulting
 
Mastering the Unannounced Regulatory Inspection
Mastering the Unannounced Regulatory InspectionMastering the Unannounced Regulatory Inspection
Mastering the Unannounced Regulatory Inspection
 
Arihant handbook biology for class 11 .pdf
Arihant handbook biology for class 11 .pdfArihant handbook biology for class 11 .pdf
Arihant handbook biology for class 11 .pdf
 
MENTAL STATUS EXAMINATION format.docx
MENTAL     STATUS EXAMINATION format.docxMENTAL     STATUS EXAMINATION format.docx
MENTAL STATUS EXAMINATION format.docx
 
Presentation by Andreas Schleicher Tackling the School Absenteeism Crisis 30 ...
Presentation by Andreas Schleicher Tackling the School Absenteeism Crisis 30 ...Presentation by Andreas Schleicher Tackling the School Absenteeism Crisis 30 ...
Presentation by Andreas Schleicher Tackling the School Absenteeism Crisis 30 ...
 
SOCIAL AND HISTORICAL CONTEXT - LFTVD.pptx
SOCIAL AND HISTORICAL CONTEXT - LFTVD.pptxSOCIAL AND HISTORICAL CONTEXT - LFTVD.pptx
SOCIAL AND HISTORICAL CONTEXT - LFTVD.pptx
 
Model Call Girl in Bikash Puri Delhi reach out to us at 🔝9953056974🔝
Model Call Girl in Bikash Puri  Delhi reach out to us at 🔝9953056974🔝Model Call Girl in Bikash Puri  Delhi reach out to us at 🔝9953056974🔝
Model Call Girl in Bikash Puri Delhi reach out to us at 🔝9953056974🔝
 
BASLIQ CURRENT LOOKBOOK LOOKBOOK(1) (1).pdf
BASLIQ CURRENT LOOKBOOK  LOOKBOOK(1) (1).pdfBASLIQ CURRENT LOOKBOOK  LOOKBOOK(1) (1).pdf
BASLIQ CURRENT LOOKBOOK LOOKBOOK(1) (1).pdf
 
Paris 2024 Olympic Geographies - an activity
Paris 2024 Olympic Geographies - an activityParis 2024 Olympic Geographies - an activity
Paris 2024 Olympic Geographies - an activity
 

Self Taught Learning

  • 1. Self-taught learning: Transfer learning from unlabelled data Authors: Rajat Raina et al. @Stanford University Presenter: Shao-Chuan Wang Some slides are from author’s work
  • 2. Self-taught learning: Transfer learning from unlabelled data Objective: Use the unlabelled data to improve the performance on a classification task. Kea Ideas: Relax the assumption about the unlabelled data. Use unlabelled data to learn the best representation (dictionary)
  • 3.
  • 6. Next: Self-taught learning?Cars Motorcycles Cars Motorcycles Cars Motorcycles Bus Tractor Aircraft Helicopter Motorcycle Car Natural scenes
  • 7. Self-taught Learning Labeled examples: Unlabeled examples: The unlabeled and labeled data: Need not share labels y. Need not share a generative distribution. Advantage: Such unlabeled data is often easy to obtain.
  • 8. Self-taught learning Use unlabelled data to learn the best representation.
  • 9. Learning the structure Sparse coding: learning the dictionary Encoding (L1-regularized least square problem) Least angle regression (Efron et al. 2004) (very fast!!) Feature-sign search (Honglak Lee et al. 2006) Coordinate descent (Friedman et al. 2008) Update dictionary(L2-constrained least square problem) K-SVD (Aharon et al. 2006) Online dictionary learning (Mairal et al. 2009) Reconstruction error Sparsity penalty
  • 10. Self-taught learning: flow K-SVD, online dictionary learning
 Learning algorithm Unlabelled data Dictionary Encoded Training features Encoding SVM with Fisher kernel Training data Encoded Testing features Testing data The Lasso Predicted Labels
  • 11. SVM with Fisher kernel Fisher kernel In Bayesian view, Laplace prior
  • 12. Example atoms Natural images. “edges” “strokes” Handwritten characters. Sparse representation gives a higher level representation
  • 13. Result: Character recognition Digits Handwritten English English font Handwritten English classification (20 labeled images per handwritten character) Bases learnt on digits English font classification (20 labeled images per font character) Bases learnt on handwritten English 8.2% error reduction 2.8% error reduction
  • 14. Image classification Other reported results: Fei-Fei et al, 2004: 16% Berg et al., 2005: 17% Holub et al., 2005: 40% Serre et al., 2005: 35% Berg et al, 2005: 48% Zhang et al., 2006: 59% Lazebnik et al., 2006: 56% (15 labeled images per class) 36.0% error reduction Self-taught Learning
  • 15. Discussion Why sparse coding? Can this idea (learning structure from unlabelled image) be applied to other encoding schemes? Probably because the nonlinear coding from x to alpha? Emulation of “end-stopping” phenomenon. (A feature is maximally activated by edges of only a specific orientation and length)
  • 16.
  • 17. Use unlabeled data to discover features.
  • 18. Use sparse coding to construct an easy-to-classify, “higher-level” representation.Unlabeled images = 0.8 * + 0.3 * + 0.5 *