Deep neural networks

S
Deep Learning and Application in
       Neural Networks


   Hugo Larochelle    Geoffrey Hinton
    Yoshua Bengio       Andrew Ng.
  Jerome Louradour   Andrew L. Nelson
    Pascal Lamblin   R. Salskhutdinov
Biological Neurons




                                        Terminal Branches
                     Dendrites
                                             of Axon




                                 Axon
Artificial Neural Networks (ANN)

                                Terminal Branches
             Dendrites
                                     of Axon
  x1
       w1
  x2
       w2
  x3   w3




                         Axon

       wn
  xn




                                Slide credit : Andrew L. Nelson
Layered Networks




Input nodes                                        Output nodes
                       Hidden nodes
        Connections

              Output : yi   f ( wi1x1 wi2x 2 wi3x 3  wim x m )
                            f(       wij x j )
                                 j
Neural network application
• ALVINN drives 70 mph on highways




                                     Slide credit : T.Michell
Neural network application




                      Slide credit : Andraw Ng.
The simplest model- the Perceptron
  •The Perceptron was introduced in 1957 by
  Frank Rosenblatt.



Perceptron:                              -                             D0


                                        d
                                                                       D1



Activation                                                             D2
functions:



Learning:                                     Input     Output
                                              Layer               Destinations
                                                        Layer
              Update




                                                      Slide credit : Geoffrey Hinton
Second generation neural networks (~1985)
                      Compare outputs with
Back-propagate        correct answer to get
error signal to get   error signal
derivatives for
learning                                             outputs




                                                            hidden
                                                            layers


                                                    input vector

                                              Slide credit : Geoffrey Hinton
BP-algorithm
                                       1
                                       .5

                                       0
                                            -5   0   5
 Activations




                       errors




The error:
                       Update




Update
Weights:
                                Slide credit : Geoffrey Hinton
Back Propagation
 Advantages
 • Multi layer Perceptron network can be
 trained by the back propagation algorithm
 to perform any mapping between the
 input and the output.

What is wrong with back-propagation?
 •It requires labeled training data.
       Almost all data is unlabeled.
 •The learning time does not scale well
       It is very slow in networks with
       multiple hidden layers.
 •It can get stuck in poor local optima.




                                             Slide credit : Geoffrey Hinton
Why Deep multi-layered neural network

              object models




              object parts
              (combination
              of edges)



             edges




         pixels
Deep neural networks
Deep Neural Networks
• Standard learning strategy
  – Randomly initializing the weights of the network
  – Applying gradient descent using backpropagation
• But, backpropagation does not work well
  (if randomly initialized)
  – Deep networks trained with back-propagation
    (without unsupervised pre-train) perform worse
    than shallow networks
  – ANN have limited to one or two layers
Deep neural networks
Deep neural networks
Deep neural networks
Unsupervised greedy layer-wise
     training procedure.
Deep neural networks
Deep neural networks
Deep neural networks
Deep neural networks
Deep neural networks
Deep neural networks
Deep neural networks
Deep neural networks
Deep neural networks
Layer-Local Unsupervised Learning
• Restricted Boltzmann Machine (SRBM) (Hinton et al,
    NC’2006)
•   Auto-encoders (Bengio et al, NIPS’2006)
•   Sparse auto-encoders (Ranzato et al, NIPS’2006)
•   Kernel PCA (Erhan 2008)
•   Denoising auto-encoders (Vincent et al, ICML’2008)
•   Unsupervised embedding (Weston et al, ICML’2008)
•   Slow features (Mohabi et al, ICML’2009, Bergstra & Bengio
    NIPS’2009)
Experiments
• MNIST data set
  – A benchmark for handwritten digit recognition
  – The number of classes is 10 (corresponding to the
    digits from 0 to 9)
  – The inputs were scaled between 0 and 1




                                Exploring Strategies for Training Deep Neural Networks.
 Hugo Larochelle, Yoshua Bengio, Jérôme Louradour, Pascal Lamblin; 10(Jan):1--40, 2009.
Classification error on MNIST training
Experiments
• Histograms presenting the test errors obtained on MNIST
  using models trained with or without pre-training.




         The difficulty of training deep architectures and the effect of unsupervised pre-training.
       Dumitru Erhan, Pierre-Antoine Manzagol, Yoshua Bengio, Samy Bengio, and Pascal Vincent;
                                                                              pages 153-160, 2009.
Experiments
• NORB data set
  – 5 object categories, 5 difference objects within
    each category.
• Classification error rate
  – DBM : 10.8 %
  – SVM : 11.6 %
  – Logistic Regression : 22.5 %
  – KNN : 18.4 %


                               Efficient Learning of Deep Boltzmann Machines.
                                        Ruslan Salakhutdinov, Hugo Larochelle ;
                                                   JMLR W&CP 9:693-700, 2010.
Conclusion
• Deep Learning : powerful arguments &
  generalization priciples
• Unsupervised Feature Learning is crucial
  many new algorithms and applications in
  recent years
• Deep Learning suited for multi-task learning,
  domain adaptation and semi-learning with
  few labels
Application
• Classification (Bengio et al., 2007; Ranzato et al., 2007b;
    Larochelle et al., 2007; Ranzato et al.,2008)
• Regression (Salakhutdinov and Hinton, 2008)
• Dimensionality Reduction (Hinton Salakhutdinov, 2006;
    Salakhutdinov and Hinton, 2007)
•   Modeling Textures (Osindero and Hinton, 2008)
•   Information Retrieval (Salakhutdinov and Hinton, 2007)
•   Robotics (Hadsell et al., 2008)
•   Natural Language Processing (Collobert and Weston, 2008;
    Weston et al., 2008)
• Collaborative Filtering (Salakhutdinov et al., 2007)
Recent Deep Learning Highlights
Recent Deep Learning Highlights
• Google Goggles uses Stacked Sparse Auto Encoders
  (Hartmut Neven @ ICML 2011)
• The monograph or review paper Learning Deep Architectures for AI
  (Foundations & Trends in Machine Learning, 2009).
• Exploring Strategies for Training Deep Neural Networks, Hugo
  Larochelle, Yoshua Bengio, Jerome Louradour and Pascal Lamblin
  in: The Journal of Machine Learning Research, pages 1-40, 2009.
• The LISA publications database contains a deep architectures
  category. http://www.iro.umontreal.ca/~lisa/twiki/bin/view.cgi/Public/
  ReadingOnDeepNetworks
• Deep Machine Learning – A New Frontier in Artificial Intelligence
  Research – a survey paper by Itamar Arel, Derek C. Rose, and
  Thomas P. Karnowski.
Thank You
1 von 36

Recomendados

Introduction to Deep Learning von
Introduction to Deep LearningIntroduction to Deep Learning
Introduction to Deep LearningOswald Campesato
3.8K views66 Folien
Back propagation von
Back propagationBack propagation
Back propagationNagarajan
66.7K views44 Folien
Deep Learning With Neural Networks von
Deep Learning With Neural NetworksDeep Learning With Neural Networks
Deep Learning With Neural NetworksAniket Maurya
3.1K views14 Folien
Deep learning von
Deep learningDeep learning
Deep learningRatnakar Pandey
11.5K views21 Folien
Deep learning - A Visual Introduction von
Deep learning - A Visual IntroductionDeep learning - A Visual Introduction
Deep learning - A Visual IntroductionLukas Masuch
57.5K views53 Folien
Deep Learning - Convolutional Neural Networks von
Deep Learning - Convolutional Neural NetworksDeep Learning - Convolutional Neural Networks
Deep Learning - Convolutional Neural NetworksChristian Perone
71.4K views86 Folien

Más contenido relacionado

Was ist angesagt?

Training Neural Networks von
Training Neural NetworksTraining Neural Networks
Training Neural NetworksDatabricks
3.6K views57 Folien
Artifical Neural Network and its applications von
Artifical Neural Network and its applicationsArtifical Neural Network and its applications
Artifical Neural Network and its applicationsSangeeta Tiwari
1.6K views31 Folien
Convolutional Neural Networks (CNN) von
Convolutional Neural Networks (CNN)Convolutional Neural Networks (CNN)
Convolutional Neural Networks (CNN)Gaurav Mittal
58.5K views70 Folien
Deep learning presentation von
Deep learning presentationDeep learning presentation
Deep learning presentationTunde Ajose-Ismail
33.4K views79 Folien
Neural Networks: Multilayer Perceptron von
Neural Networks: Multilayer PerceptronNeural Networks: Multilayer Perceptron
Neural Networks: Multilayer PerceptronMostafa G. M. Mostafa
8.7K views42 Folien
Deep learning von
Deep learning Deep learning
Deep learning Rajgupta258
1.6K views20 Folien

Was ist angesagt?(20)

Training Neural Networks von Databricks
Training Neural NetworksTraining Neural Networks
Training Neural Networks
Databricks3.6K views
Artifical Neural Network and its applications von Sangeeta Tiwari
Artifical Neural Network and its applicationsArtifical Neural Network and its applications
Artifical Neural Network and its applications
Sangeeta Tiwari1.6K views
Convolutional Neural Networks (CNN) von Gaurav Mittal
Convolutional Neural Networks (CNN)Convolutional Neural Networks (CNN)
Convolutional Neural Networks (CNN)
Gaurav Mittal58.5K views
Deep learning von Rajgupta258
Deep learning Deep learning
Deep learning
Rajgupta2581.6K views
Deep learning - Conceptual understanding and applications von Buhwan Jeong
Deep learning - Conceptual understanding and applicationsDeep learning - Conceptual understanding and applications
Deep learning - Conceptual understanding and applications
Buhwan Jeong74.4K views
Convolutional neural network von Yan Xu
Convolutional neural network Convolutional neural network
Convolutional neural network
Yan Xu5.3K views
Introduction to Deep learning von leopauly
Introduction to Deep learningIntroduction to Deep learning
Introduction to Deep learning
leopauly1.7K views
Machine Learning: Introduction to Neural Networks von Francesco Collova'
Machine Learning: Introduction to Neural NetworksMachine Learning: Introduction to Neural Networks
Machine Learning: Introduction to Neural Networks
Francesco Collova'16.1K views
Deep Learning Explained von Melanie Swan
Deep Learning ExplainedDeep Learning Explained
Deep Learning Explained
Melanie Swan41.4K views
Artificial Neural Network seminar presentation using ppt. von Mohd Faiz
Artificial Neural Network seminar presentation using ppt.Artificial Neural Network seminar presentation using ppt.
Artificial Neural Network seminar presentation using ppt.
Mohd Faiz5.7K views
Recurrent Neural Network (RNN) | RNN LSTM Tutorial | Deep Learning Course | S... von Simplilearn
Recurrent Neural Network (RNN) | RNN LSTM Tutorial | Deep Learning Course | S...Recurrent Neural Network (RNN) | RNN LSTM Tutorial | Deep Learning Course | S...
Recurrent Neural Network (RNN) | RNN LSTM Tutorial | Deep Learning Course | S...
Simplilearn18.8K views
An introduction to Deep Learning von Julien SIMON
An introduction to Deep LearningAn introduction to Deep Learning
An introduction to Deep Learning
Julien SIMON4.3K views
Introduction of Deep Learning von Myungjin Lee
Introduction of Deep LearningIntroduction of Deep Learning
Introduction of Deep Learning
Myungjin Lee6K views
Deep Learning: Application & Opportunity von iTrain
Deep Learning: Application & OpportunityDeep Learning: Application & Opportunity
Deep Learning: Application & Opportunity
iTrain1.6K views

Destacado

Deep Learning through Examples von
Deep Learning through ExamplesDeep Learning through Examples
Deep Learning through ExamplesSri Ambati
150.6K views47 Folien
Bắt đầu học data science von
Bắt đầu học data scienceBắt đầu học data science
Bắt đầu học data scienceHong Ong
88K views46 Folien
Bắt đầu nghiên cứu Big Data von
Bắt đầu nghiên cứu Big DataBắt đầu nghiên cứu Big Data
Bắt đầu nghiên cứu Big DataHong Ong
98.5K views36 Folien
Hands-on Deep Learning in Python von
Hands-on Deep Learning in PythonHands-on Deep Learning in Python
Hands-on Deep Learning in PythonImry Kissos
30.7K views65 Folien
Hadoop and Machine Learning von
Hadoop and Machine LearningHadoop and Machine Learning
Hadoop and Machine Learningjoshwills
53.1K views26 Folien
A Statistician's View on Big Data and Data Science (Version 1) von
A Statistician's View on Big Data and Data Science (Version 1)A Statistician's View on Big Data and Data Science (Version 1)
A Statistician's View on Big Data and Data Science (Version 1)Prof. Dr. Diego Kuonen
95.4K views40 Folien

Destacado(20)

Deep Learning through Examples von Sri Ambati
Deep Learning through ExamplesDeep Learning through Examples
Deep Learning through Examples
Sri Ambati150.6K views
Bắt đầu học data science von Hong Ong
Bắt đầu học data scienceBắt đầu học data science
Bắt đầu học data science
Hong Ong88K views
Bắt đầu nghiên cứu Big Data von Hong Ong
Bắt đầu nghiên cứu Big DataBắt đầu nghiên cứu Big Data
Bắt đầu nghiên cứu Big Data
Hong Ong98.5K views
Hands-on Deep Learning in Python von Imry Kissos
Hands-on Deep Learning in PythonHands-on Deep Learning in Python
Hands-on Deep Learning in Python
Imry Kissos30.7K views
Hadoop and Machine Learning von joshwills
Hadoop and Machine LearningHadoop and Machine Learning
Hadoop and Machine Learning
joshwills53.1K views
A Statistician's View on Big Data and Data Science (Version 1) von Prof. Dr. Diego Kuonen
A Statistician's View on Big Data and Data Science (Version 1)A Statistician's View on Big Data and Data Science (Version 1)
A Statistician's View on Big Data and Data Science (Version 1)
Data By The People, For The People von Daniel Tunkelang
Data By The People, For The PeopleData By The People, For The People
Data By The People, For The People
Daniel Tunkelang83.1K views
10 Lessons Learned from Building Machine Learning Systems von Xavier Amatriain
10 Lessons Learned from Building Machine Learning Systems10 Lessons Learned from Building Machine Learning Systems
10 Lessons Learned from Building Machine Learning Systems
Xavier Amatriain378.5K views
Big Data [sorry] & Data Science: What Does a Data Scientist Do? von Data Science London
Big Data [sorry] & Data Science: What Does a Data Scientist Do?Big Data [sorry] & Data Science: What Does a Data Scientist Do?
Big Data [sorry] & Data Science: What Does a Data Scientist Do?
Data Science London147.9K views
How to Become a Data Scientist von ryanorban
How to Become a Data ScientistHow to Become a Data Scientist
How to Become a Data Scientist
ryanorban197.6K views
A tutorial on deep learning at icml 2013 von Philip Zheng
A tutorial on deep learning at icml 2013A tutorial on deep learning at icml 2013
A tutorial on deep learning at icml 2013
Philip Zheng45.4K views
Deep Learning for Natural Language Processing von Devashish Shanker
Deep Learning for Natural Language ProcessingDeep Learning for Natural Language Processing
Deep Learning for Natural Language Processing
Devashish Shanker32K views
Introduction to Mahout and Machine Learning von Varad Meru
Introduction to Mahout and Machine LearningIntroduction to Mahout and Machine Learning
Introduction to Mahout and Machine Learning
Varad Meru77K views
An Introduction to Supervised Machine Learning and Pattern Classification: Th... von Sebastian Raschka
An Introduction to Supervised Machine Learning and Pattern Classification: Th...An Introduction to Supervised Machine Learning and Pattern Classification: Th...
An Introduction to Supervised Machine Learning and Pattern Classification: Th...
Sebastian Raschka101.9K views
Machine Learning and Data Mining: 12 Classification Rules von Pier Luca Lanzi
Machine Learning and Data Mining: 12 Classification RulesMachine Learning and Data Mining: 12 Classification Rules
Machine Learning and Data Mining: 12 Classification Rules
Pier Luca Lanzi56.2K views
Myths and Mathemagical Superpowers of Data Scientists von David Pittman
Myths and Mathemagical Superpowers of Data ScientistsMyths and Mathemagical Superpowers of Data Scientists
Myths and Mathemagical Superpowers of Data Scientists
David Pittman96.7K views
Tutorial on Deep learning and Applications von NhatHai Phan
Tutorial on Deep learning and ApplicationsTutorial on Deep learning and Applications
Tutorial on Deep learning and Applications
NhatHai Phan28.9K views
Tips for data science competitions von Owen Zhang
Tips for data science competitionsTips for data science competitions
Tips for data science competitions
Owen Zhang85.6K views
Introduction to Big Data/Machine Learning von Lars Marius Garshol
Introduction to Big Data/Machine LearningIntroduction to Big Data/Machine Learning
Introduction to Big Data/Machine Learning
Lars Marius Garshol306.6K views

Similar a Deep neural networks

introduction to deeplearning von
introduction to deeplearningintroduction to deeplearning
introduction to deeplearningEyad Alshami
902 views36 Folien
AI&BigData Lab. Артем Чернодуб "Распознавание изображений методом Lazy Deep ... von
AI&BigData Lab. Артем Чернодуб  "Распознавание изображений методом Lazy Deep ...AI&BigData Lab. Артем Чернодуб  "Распознавание изображений методом Lazy Deep ...
AI&BigData Lab. Артем Чернодуб "Распознавание изображений методом Lazy Deep ...GeeksLab Odessa
3K views55 Folien
AI&BigData Lab 2016. Артем Чернодуб: Обучение глубоких, очень глубоких и реку... von
AI&BigData Lab 2016. Артем Чернодуб: Обучение глубоких, очень глубоких и реку...AI&BigData Lab 2016. Артем Чернодуб: Обучение глубоких, очень глубоких и реку...
AI&BigData Lab 2016. Артем Чернодуб: Обучение глубоких, очень глубоких и реку...GeeksLab Odessa
1.1K views47 Folien
Fundamental of deep learning von
Fundamental of deep learningFundamental of deep learning
Fundamental of deep learningStanley Wang
1.3K views62 Folien
What's Wrong With Deep Learning? von
What's Wrong With Deep Learning?What's Wrong With Deep Learning?
What's Wrong With Deep Learning?Philip Zheng
2.1K views157 Folien
Evolution of Deep Learning and new advancements von
Evolution of Deep Learning and new advancementsEvolution of Deep Learning and new advancements
Evolution of Deep Learning and new advancementsChitta Ranjan
138 views69 Folien

Similar a Deep neural networks(20)

introduction to deeplearning von Eyad Alshami
introduction to deeplearningintroduction to deeplearning
introduction to deeplearning
Eyad Alshami902 views
AI&BigData Lab. Артем Чернодуб "Распознавание изображений методом Lazy Deep ... von GeeksLab Odessa
AI&BigData Lab. Артем Чернодуб  "Распознавание изображений методом Lazy Deep ...AI&BigData Lab. Артем Чернодуб  "Распознавание изображений методом Lazy Deep ...
AI&BigData Lab. Артем Чернодуб "Распознавание изображений методом Lazy Deep ...
GeeksLab Odessa3K views
AI&BigData Lab 2016. Артем Чернодуб: Обучение глубоких, очень глубоких и реку... von GeeksLab Odessa
AI&BigData Lab 2016. Артем Чернодуб: Обучение глубоких, очень глубоких и реку...AI&BigData Lab 2016. Артем Чернодуб: Обучение глубоких, очень глубоких и реку...
AI&BigData Lab 2016. Артем Чернодуб: Обучение глубоких, очень глубоких и реку...
GeeksLab Odessa1.1K views
Fundamental of deep learning von Stanley Wang
Fundamental of deep learningFundamental of deep learning
Fundamental of deep learning
Stanley Wang1.3K views
What's Wrong With Deep Learning? von Philip Zheng
What's Wrong With Deep Learning?What's Wrong With Deep Learning?
What's Wrong With Deep Learning?
Philip Zheng2.1K views
Evolution of Deep Learning and new advancements von Chitta Ranjan
Evolution of Deep Learning and new advancementsEvolution of Deep Learning and new advancements
Evolution of Deep Learning and new advancements
Chitta Ranjan138 views
Deep learning from a novice perspective von Anirban Santara
Deep learning from a novice perspectiveDeep learning from a novice perspective
Deep learning from a novice perspective
Anirban Santara1.2K views
An Introduction to Deep Learning von Poo Kuan Hoong
An Introduction to Deep LearningAn Introduction to Deep Learning
An Introduction to Deep Learning
Poo Kuan Hoong11.8K views
Big Data Malaysia - A Primer on Deep Learning von Poo Kuan Hoong
Big Data Malaysia - A Primer on Deep LearningBig Data Malaysia - A Primer on Deep Learning
Big Data Malaysia - A Primer on Deep Learning
Poo Kuan Hoong696 views
MDEC Data Matters Series: machine learning and Deep Learning, A Primer von Poo Kuan Hoong
MDEC Data Matters Series: machine learning and Deep Learning, A PrimerMDEC Data Matters Series: machine learning and Deep Learning, A Primer
MDEC Data Matters Series: machine learning and Deep Learning, A Primer
Poo Kuan Hoong1.1K views
Details of Lazy Deep Learning for Images Recognition in ZZ Photo app von PAY2 YOU
Details of Lazy Deep Learning for Images Recognition in ZZ Photo appDetails of Lazy Deep Learning for Images Recognition in ZZ Photo app
Details of Lazy Deep Learning for Images Recognition in ZZ Photo app
PAY2 YOU9.3K views
Autoencoders for image_classification von Cenk Bircanoğlu
Autoencoders for image_classificationAutoencoders for image_classification
Autoencoders for image_classification
Cenk Bircanoğlu4.9K views
DEF CON 24 - Clarence Chio - machine duping 101 von Felipe Prado
DEF CON 24 - Clarence Chio - machine duping 101DEF CON 24 - Clarence Chio - machine duping 101
DEF CON 24 - Clarence Chio - machine duping 101
Felipe Prado70 views
Deep Learning - 인공지능 기계학습의 새로운 트랜드 :김인중 von datasciencekorea
Deep Learning - 인공지능 기계학습의 새로운 트랜드 :김인중Deep Learning - 인공지능 기계학습의 새로운 트랜드 :김인중
Deep Learning - 인공지능 기계학습의 새로운 트랜드 :김인중
datasciencekorea8.6K views
2010 deep learning and unsupervised feature learning von Van Thanh
2010 deep learning and unsupervised feature learning2010 deep learning and unsupervised feature learning
2010 deep learning and unsupervised feature learning
Van Thanh1.7K views
Convolutional neural networks 이론과 응용 von 홍배 김
Convolutional neural networks 이론과 응용Convolutional neural networks 이론과 응용
Convolutional neural networks 이론과 응용
홍배 김2.2K views

Último

Career Building in AI - Technologies, Trends and Opportunities von
Career Building in AI - Technologies, Trends and OpportunitiesCareer Building in AI - Technologies, Trends and Opportunities
Career Building in AI - Technologies, Trends and OpportunitiesWebStackAcademy
41 views44 Folien
Papal.pdf von
Papal.pdfPapal.pdf
Papal.pdfMariaKenney3
57 views24 Folien
INT-244 Topic 6b Confucianism von
INT-244 Topic 6b ConfucianismINT-244 Topic 6b Confucianism
INT-244 Topic 6b ConfucianismS Meyer
44 views77 Folien
Payment Integration using Braintree Connector | MuleSoft Mysore Meetup #37 von
Payment Integration using Braintree Connector | MuleSoft Mysore Meetup #37Payment Integration using Braintree Connector | MuleSoft Mysore Meetup #37
Payment Integration using Braintree Connector | MuleSoft Mysore Meetup #37MysoreMuleSoftMeetup
44 views17 Folien
BÀI TẬP BỔ TRỢ TIẾNG ANH 11 THEO ĐƠN VỊ BÀI HỌC - CẢ NĂM - CÓ FILE NGHE (FRIE... von
BÀI TẬP BỔ TRỢ TIẾNG ANH 11 THEO ĐƠN VỊ BÀI HỌC - CẢ NĂM - CÓ FILE NGHE (FRIE...BÀI TẬP BỔ TRỢ TIẾNG ANH 11 THEO ĐƠN VỊ BÀI HỌC - CẢ NĂM - CÓ FILE NGHE (FRIE...
BÀI TẬP BỔ TRỢ TIẾNG ANH 11 THEO ĐƠN VỊ BÀI HỌC - CẢ NĂM - CÓ FILE NGHE (FRIE...Nguyen Thanh Tu Collection
71 views91 Folien
JQUERY.pdf von
JQUERY.pdfJQUERY.pdf
JQUERY.pdfArthyR3
103 views22 Folien

Último(20)

Career Building in AI - Technologies, Trends and Opportunities von WebStackAcademy
Career Building in AI - Technologies, Trends and OpportunitiesCareer Building in AI - Technologies, Trends and Opportunities
Career Building in AI - Technologies, Trends and Opportunities
WebStackAcademy41 views
INT-244 Topic 6b Confucianism von S Meyer
INT-244 Topic 6b ConfucianismINT-244 Topic 6b Confucianism
INT-244 Topic 6b Confucianism
S Meyer44 views
Payment Integration using Braintree Connector | MuleSoft Mysore Meetup #37 von MysoreMuleSoftMeetup
Payment Integration using Braintree Connector | MuleSoft Mysore Meetup #37Payment Integration using Braintree Connector | MuleSoft Mysore Meetup #37
Payment Integration using Braintree Connector | MuleSoft Mysore Meetup #37
BÀI TẬP BỔ TRỢ TIẾNG ANH 11 THEO ĐƠN VỊ BÀI HỌC - CẢ NĂM - CÓ FILE NGHE (FRIE... von Nguyen Thanh Tu Collection
BÀI TẬP BỔ TRỢ TIẾNG ANH 11 THEO ĐƠN VỊ BÀI HỌC - CẢ NĂM - CÓ FILE NGHE (FRIE...BÀI TẬP BỔ TRỢ TIẾNG ANH 11 THEO ĐƠN VỊ BÀI HỌC - CẢ NĂM - CÓ FILE NGHE (FRIE...
BÀI TẬP BỔ TRỢ TIẾNG ANH 11 THEO ĐƠN VỊ BÀI HỌC - CẢ NĂM - CÓ FILE NGHE (FRIE...
JQUERY.pdf von ArthyR3
JQUERY.pdfJQUERY.pdf
JQUERY.pdf
ArthyR3103 views
Create a Structure in VBNet.pptx von Breach_P
Create a Structure in VBNet.pptxCreate a Structure in VBNet.pptx
Create a Structure in VBNet.pptx
Breach_P82 views
The Future of Micro-credentials: Is Small Really Beautiful? von Mark Brown
The Future of Micro-credentials:  Is Small Really Beautiful?The Future of Micro-credentials:  Is Small Really Beautiful?
The Future of Micro-credentials: Is Small Really Beautiful?
Mark Brown48 views
NodeJS and ExpressJS.pdf von ArthyR3
NodeJS and ExpressJS.pdfNodeJS and ExpressJS.pdf
NodeJS and ExpressJS.pdf
ArthyR347 views
Pharmaceutical Analysis PPT (BP 102T) von yakshpharmacy009
Pharmaceutical Analysis PPT (BP 102T) Pharmaceutical Analysis PPT (BP 102T)
Pharmaceutical Analysis PPT (BP 102T)
yakshpharmacy009101 views
CUNY IT Picciano.pptx von apicciano
CUNY IT Picciano.pptxCUNY IT Picciano.pptx
CUNY IT Picciano.pptx
apicciano60 views
JRN 362 - Lecture Twenty-Three (Epilogue) von Rich Hanley
JRN 362 - Lecture Twenty-Three (Epilogue)JRN 362 - Lecture Twenty-Three (Epilogue)
JRN 362 - Lecture Twenty-Three (Epilogue)
Rich Hanley41 views

Deep neural networks

  • 1. Deep Learning and Application in Neural Networks Hugo Larochelle Geoffrey Hinton Yoshua Bengio Andrew Ng. Jerome Louradour Andrew L. Nelson Pascal Lamblin R. Salskhutdinov
  • 2. Biological Neurons Terminal Branches Dendrites of Axon Axon
  • 3. Artificial Neural Networks (ANN) Terminal Branches Dendrites of Axon x1 w1 x2 w2 x3 w3 Axon wn xn Slide credit : Andrew L. Nelson
  • 4. Layered Networks Input nodes Output nodes Hidden nodes Connections Output : yi f ( wi1x1 wi2x 2 wi3x 3  wim x m ) f( wij x j ) j
  • 5. Neural network application • ALVINN drives 70 mph on highways Slide credit : T.Michell
  • 6. Neural network application Slide credit : Andraw Ng.
  • 7. The simplest model- the Perceptron •The Perceptron was introduced in 1957 by Frank Rosenblatt. Perceptron: - D0 d D1 Activation D2 functions: Learning: Input Output Layer Destinations Layer Update Slide credit : Geoffrey Hinton
  • 8. Second generation neural networks (~1985) Compare outputs with Back-propagate correct answer to get error signal to get error signal derivatives for learning outputs hidden layers input vector Slide credit : Geoffrey Hinton
  • 9. BP-algorithm 1 .5 0 -5 0 5 Activations errors The error: Update Update Weights: Slide credit : Geoffrey Hinton
  • 10. Back Propagation Advantages • Multi layer Perceptron network can be trained by the back propagation algorithm to perform any mapping between the input and the output. What is wrong with back-propagation? •It requires labeled training data. Almost all data is unlabeled. •The learning time does not scale well It is very slow in networks with multiple hidden layers. •It can get stuck in poor local optima. Slide credit : Geoffrey Hinton
  • 11. Why Deep multi-layered neural network object models object parts (combination of edges) edges pixels
  • 13. Deep Neural Networks • Standard learning strategy – Randomly initializing the weights of the network – Applying gradient descent using backpropagation • But, backpropagation does not work well (if randomly initialized) – Deep networks trained with back-propagation (without unsupervised pre-train) perform worse than shallow networks – ANN have limited to one or two layers
  • 17. Unsupervised greedy layer-wise training procedure.
  • 27. Layer-Local Unsupervised Learning • Restricted Boltzmann Machine (SRBM) (Hinton et al, NC’2006) • Auto-encoders (Bengio et al, NIPS’2006) • Sparse auto-encoders (Ranzato et al, NIPS’2006) • Kernel PCA (Erhan 2008) • Denoising auto-encoders (Vincent et al, ICML’2008) • Unsupervised embedding (Weston et al, ICML’2008) • Slow features (Mohabi et al, ICML’2009, Bergstra & Bengio NIPS’2009)
  • 28. Experiments • MNIST data set – A benchmark for handwritten digit recognition – The number of classes is 10 (corresponding to the digits from 0 to 9) – The inputs were scaled between 0 and 1 Exploring Strategies for Training Deep Neural Networks. Hugo Larochelle, Yoshua Bengio, Jérôme Louradour, Pascal Lamblin; 10(Jan):1--40, 2009.
  • 29. Classification error on MNIST training
  • 30. Experiments • Histograms presenting the test errors obtained on MNIST using models trained with or without pre-training. The difficulty of training deep architectures and the effect of unsupervised pre-training. Dumitru Erhan, Pierre-Antoine Manzagol, Yoshua Bengio, Samy Bengio, and Pascal Vincent; pages 153-160, 2009.
  • 31. Experiments • NORB data set – 5 object categories, 5 difference objects within each category. • Classification error rate – DBM : 10.8 % – SVM : 11.6 % – Logistic Regression : 22.5 % – KNN : 18.4 % Efficient Learning of Deep Boltzmann Machines. Ruslan Salakhutdinov, Hugo Larochelle ; JMLR W&CP 9:693-700, 2010.
  • 32. Conclusion • Deep Learning : powerful arguments & generalization priciples • Unsupervised Feature Learning is crucial many new algorithms and applications in recent years • Deep Learning suited for multi-task learning, domain adaptation and semi-learning with few labels
  • 33. Application • Classification (Bengio et al., 2007; Ranzato et al., 2007b; Larochelle et al., 2007; Ranzato et al.,2008) • Regression (Salakhutdinov and Hinton, 2008) • Dimensionality Reduction (Hinton Salakhutdinov, 2006; Salakhutdinov and Hinton, 2007) • Modeling Textures (Osindero and Hinton, 2008) • Information Retrieval (Salakhutdinov and Hinton, 2007) • Robotics (Hadsell et al., 2008) • Natural Language Processing (Collobert and Weston, 2008; Weston et al., 2008) • Collaborative Filtering (Salakhutdinov et al., 2007)
  • 34. Recent Deep Learning Highlights
  • 35. Recent Deep Learning Highlights • Google Goggles uses Stacked Sparse Auto Encoders (Hartmut Neven @ ICML 2011) • The monograph or review paper Learning Deep Architectures for AI (Foundations & Trends in Machine Learning, 2009). • Exploring Strategies for Training Deep Neural Networks, Hugo Larochelle, Yoshua Bengio, Jerome Louradour and Pascal Lamblin in: The Journal of Machine Learning Research, pages 1-40, 2009. • The LISA publications database contains a deep architectures category. http://www.iro.umontreal.ca/~lisa/twiki/bin/view.cgi/Public/ ReadingOnDeepNetworks • Deep Machine Learning – A New Frontier in Artificial Intelligence Research – a survey paper by Itamar Arel, Derek C. Rose, and Thomas P. Karnowski.