SlideShare a Scribd company logo
Exploring Strategies for Training Deep
Neural Networks
By Hugo Larochelle, Yoshua Bengio,Jerome Louradour, Pascal Lamblin
By V B Wickramasinghe (148245F)
Outline
● Introduction
● Deep neural networks
● Stacked Restricted Boltzmann Machine Network
● Stacked Autoassociators Network
● Experimental results
● Conclusion
Introduction
● Training deep neural network is hard.
● This is mainly due to randomly initialized deep
architecture tend to get stuck in poor situations.
● But the ability of deep architectures to represent
complex functions is unmatched.
● This paper highlights some of the recent breakthroughs
in training deep architectures that has helped to uncover
their potential.
Deep neural networks
● Shallow networks has been proved to be inefficient in circuit theory,
boolean logic and neural networks.
● This is because some functions that can be represented using k layers is
with finite number of units takes exponential number units with k-1 layers.
● Also highly varying function can be easily represented by a number of
non-linearities stacked together.
● Another issue with shallow architectures is that they’ll require exponential
number of training examples to learn complex functions
● But as mentioned earlier training deep architectures is hard. What is the
solution?
Deep neural networks
Stacked Restricted Boltzmann Machine
Network
● RBMs represent a generative model of input.
● Train individual layers of RBMs using contrastive
divergence.
● Then stack them together so that a one layers output
representation works as input to another(A DBN).
● Hinton(2006) argues that this helps in a more complex
representation overall.
● Then the pretrained stacked framework can be trained
to for a particular task using backpropagation.
Stacked Autoassociators Network
● Like RBMs autoassociators are a type of network that when combined
helps improving input representation.
● Autoassociators are an encoding model which is trained to minimize the
reconstruction loss of input from output.
● Stacked autoassociator performs same layer wise training procedure as
DBNs.
● Reconstruction error of an autoassociator and log-likelihood of RBM are
both approximate values of convergent series of log-likelihood gradient
obtained in different ways.
Stacked Autoassociators Network
Experimental results
Experimental results
Experimental results
Experimental results
Conclusion
● DNNs are an indispensable tool for learning tasks.
● This paper presents 3 methods of optimally training DNNs,
1. pre-training one layer at a time in a greedy way.
2. using unsupervised learning at each layer in a way that preserves
information from the input and disentangles factors of variation.
3. fine-tuning the whole network with respect to the ultimate criterion of
interest.
● The experiments are sound and present clearly why deep neural networks
trained using the presented methods can help in improving learning tasks
significantly over single layer networks.

More Related Content

What's hot

On the cross domain reusability of neural modules for general video game playing
On the cross domain reusability of neural modules for general video game playingOn the cross domain reusability of neural modules for general video game playing
On the cross domain reusability of neural modules for general video game playing
Alexander Braylan
 
Practical Block-wise Neural Network Architecture Generation
Practical Block-wise Neural Network Architecture GenerationPractical Block-wise Neural Network Architecture Generation
Practical Block-wise Neural Network Architecture Generation
郁凱 黃
 
Reading group nfm - 20170312
Reading group  nfm - 20170312Reading group  nfm - 20170312
Reading group nfm - 20170312
Shuai Zhang
 
Introduction to Tree-LSTMs
Introduction to Tree-LSTMsIntroduction to Tree-LSTMs
Introduction to Tree-LSTMs
Daniel Perez
 
Introduction to CNN
Introduction to CNNIntroduction to CNN
Introduction to CNN
Shuai Zhang
 
Emnlp2015 reading festival_lstm_cws
Emnlp2015 reading festival_lstm_cwsEmnlp2015 reading festival_lstm_cws
Emnlp2015 reading festival_lstm_cws
Ace12358
 
Sequential Reptile_Inter-Task Gradient Alignment for Multilingual Learning
Sequential Reptile_Inter-Task Gradient Alignment for Multilingual LearningSequential Reptile_Inter-Task Gradient Alignment for Multilingual Learning
Sequential Reptile_Inter-Task Gradient Alignment for Multilingual Learning
MLAI2
 
Logic gates II presentation
Logic gates II presentationLogic gates II presentation
Logic gates II presentation
AhmedElazhari1
 
1D Convolutional Neural Networks for Time Series Modeling - Nathan Janos, Jef...
1D Convolutional Neural Networks for Time Series Modeling - Nathan Janos, Jef...1D Convolutional Neural Networks for Time Series Modeling - Nathan Janos, Jef...
1D Convolutional Neural Networks for Time Series Modeling - Nathan Janos, Jef...
PyData
 
RNN & LSTM: Neural Network for Sequential Data
RNN & LSTM: Neural Network for Sequential DataRNN & LSTM: Neural Network for Sequential Data
RNN & LSTM: Neural Network for Sequential Data
Yao-Chieh Hu
 
Review-image-segmentation-by-deep-learning
Review-image-segmentation-by-deep-learningReview-image-segmentation-by-deep-learning
Review-image-segmentation-by-deep-learning
Trong-An Bui
 
Neural Network Architectures
Neural Network ArchitecturesNeural Network Architectures
Neural Network Architectures
Martin Ockajak
 
SummaRuNNer: A Recurrent Neural Network based Sequence Model for Extractive ...
 SummaRuNNer: A Recurrent Neural Network based Sequence Model for Extractive ... SummaRuNNer: A Recurrent Neural Network based Sequence Model for Extractive ...
SummaRuNNer: A Recurrent Neural Network based Sequence Model for Extractive ...
Shubhangi Tandon
 
Recent Progress in RNN and NLP
Recent Progress in RNN and NLPRecent Progress in RNN and NLP
Recent Progress in RNN and NLP
hytae
 
Recurrent Neural Networks for Text Analysis
Recurrent Neural Networks for Text AnalysisRecurrent Neural Networks for Text Analysis
Recurrent Neural Networks for Text Analysis
odsc
 
PR-207: YOLOv3: An Incremental Improvement
PR-207: YOLOv3: An Incremental ImprovementPR-207: YOLOv3: An Incremental Improvement
PR-207: YOLOv3: An Incremental Improvement
Jinwon Lee
 
Deep Neural Machine Translation with Linear Associative Unit
Deep Neural Machine Translation with Linear Associative UnitDeep Neural Machine Translation with Linear Associative Unit
Deep Neural Machine Translation with Linear Associative Unit
Satoru Katsumata
 
Functional Domain Modeling
Functional Domain ModelingFunctional Domain Modeling
Functional Domain Modeling
Michal Bigos
 
Meta Dropout: Learning to Perturb Latent Features for Generalization
Meta Dropout: Learning to Perturb Latent Features for Generalization Meta Dropout: Learning to Perturb Latent Features for Generalization
Meta Dropout: Learning to Perturb Latent Features for Generalization
MLAI2
 
Using Multi-layered Feed-forward Neural Network (MLFNN) Architecture as Bidir...
Using Multi-layered Feed-forward Neural Network (MLFNN) Architecture as Bidir...Using Multi-layered Feed-forward Neural Network (MLFNN) Architecture as Bidir...
Using Multi-layered Feed-forward Neural Network (MLFNN) Architecture as Bidir...
IOSR Journals
 

What's hot (20)

On the cross domain reusability of neural modules for general video game playing
On the cross domain reusability of neural modules for general video game playingOn the cross domain reusability of neural modules for general video game playing
On the cross domain reusability of neural modules for general video game playing
 
Practical Block-wise Neural Network Architecture Generation
Practical Block-wise Neural Network Architecture GenerationPractical Block-wise Neural Network Architecture Generation
Practical Block-wise Neural Network Architecture Generation
 
Reading group nfm - 20170312
Reading group  nfm - 20170312Reading group  nfm - 20170312
Reading group nfm - 20170312
 
Introduction to Tree-LSTMs
Introduction to Tree-LSTMsIntroduction to Tree-LSTMs
Introduction to Tree-LSTMs
 
Introduction to CNN
Introduction to CNNIntroduction to CNN
Introduction to CNN
 
Emnlp2015 reading festival_lstm_cws
Emnlp2015 reading festival_lstm_cwsEmnlp2015 reading festival_lstm_cws
Emnlp2015 reading festival_lstm_cws
 
Sequential Reptile_Inter-Task Gradient Alignment for Multilingual Learning
Sequential Reptile_Inter-Task Gradient Alignment for Multilingual LearningSequential Reptile_Inter-Task Gradient Alignment for Multilingual Learning
Sequential Reptile_Inter-Task Gradient Alignment for Multilingual Learning
 
Logic gates II presentation
Logic gates II presentationLogic gates II presentation
Logic gates II presentation
 
1D Convolutional Neural Networks for Time Series Modeling - Nathan Janos, Jef...
1D Convolutional Neural Networks for Time Series Modeling - Nathan Janos, Jef...1D Convolutional Neural Networks for Time Series Modeling - Nathan Janos, Jef...
1D Convolutional Neural Networks for Time Series Modeling - Nathan Janos, Jef...
 
RNN & LSTM: Neural Network for Sequential Data
RNN & LSTM: Neural Network for Sequential DataRNN & LSTM: Neural Network for Sequential Data
RNN & LSTM: Neural Network for Sequential Data
 
Review-image-segmentation-by-deep-learning
Review-image-segmentation-by-deep-learningReview-image-segmentation-by-deep-learning
Review-image-segmentation-by-deep-learning
 
Neural Network Architectures
Neural Network ArchitecturesNeural Network Architectures
Neural Network Architectures
 
SummaRuNNer: A Recurrent Neural Network based Sequence Model for Extractive ...
 SummaRuNNer: A Recurrent Neural Network based Sequence Model for Extractive ... SummaRuNNer: A Recurrent Neural Network based Sequence Model for Extractive ...
SummaRuNNer: A Recurrent Neural Network based Sequence Model for Extractive ...
 
Recent Progress in RNN and NLP
Recent Progress in RNN and NLPRecent Progress in RNN and NLP
Recent Progress in RNN and NLP
 
Recurrent Neural Networks for Text Analysis
Recurrent Neural Networks for Text AnalysisRecurrent Neural Networks for Text Analysis
Recurrent Neural Networks for Text Analysis
 
PR-207: YOLOv3: An Incremental Improvement
PR-207: YOLOv3: An Incremental ImprovementPR-207: YOLOv3: An Incremental Improvement
PR-207: YOLOv3: An Incremental Improvement
 
Deep Neural Machine Translation with Linear Associative Unit
Deep Neural Machine Translation with Linear Associative UnitDeep Neural Machine Translation with Linear Associative Unit
Deep Neural Machine Translation with Linear Associative Unit
 
Functional Domain Modeling
Functional Domain ModelingFunctional Domain Modeling
Functional Domain Modeling
 
Meta Dropout: Learning to Perturb Latent Features for Generalization
Meta Dropout: Learning to Perturb Latent Features for Generalization Meta Dropout: Learning to Perturb Latent Features for Generalization
Meta Dropout: Learning to Perturb Latent Features for Generalization
 
Using Multi-layered Feed-forward Neural Network (MLFNN) Architecture as Bidir...
Using Multi-layered Feed-forward Neural Network (MLFNN) Architecture as Bidir...Using Multi-layered Feed-forward Neural Network (MLFNN) Architecture as Bidir...
Using Multi-layered Feed-forward Neural Network (MLFNN) Architecture as Bidir...
 

Viewers also liked

Curso auxiliar de enfermería
Curso auxiliar de enfermeríaCurso auxiliar de enfermería
Curso auxiliar de enfermería
Euroinnova Formación
 
ONG and crowdfunding case - 2012
ONG and crowdfunding case - 2012ONG and crowdfunding case - 2012
ONG and crowdfunding case - 2012
Julien Ferla
 
Asistencia de cabildo 06
Asistencia de cabildo 06Asistencia de cabildo 06
Asistencia de cabildo 06sabmpio
 
Cv B Marco Crb Rev6 Ij
Cv B Marco Crb Rev6 IjCv B Marco Crb Rev6 Ij
Cv B Marco Crb Rev6 Ijbenmaralc
 
Encontro com a escritora Rosa Duarte
Encontro com a escritora Rosa Duarte Encontro com a escritora Rosa Duarte
Encontro com a escritora Rosa Duarte António Pires
 
Magazine con alma de blues n18
Magazine con alma de blues n18Magazine con alma de blues n18
Magazine con alma de blues n18
Gustavo pollo Zungri
 
Red de rarea local2
Red de rarea local2Red de rarea local2
Red de rarea local2
Lizz Ibañez
 
EnerEscolas
EnerEscolasEnerEscolas
EnerEscolas
lpizzacalla
 
HBSL - Presentation for Educational Institutions March-2016
HBSL - Presentation for Educational Institutions March-2016HBSL - Presentation for Educational Institutions March-2016
HBSL - Presentation for Educational Institutions March-2016Afshan Siddiqui
 
Normas Insutec Virtual
Normas Insutec VirtualNormas Insutec Virtual
Normas Insutec Virtual
insutecvirtual
 
[En]ICFecc 2010 sponsor
[En]ICFecc 2010 sponsor[En]ICFecc 2010 sponsor
[En]ICFecc 2010 sponsor
ICFFrance
 
Slideshare MyElaN in French (Belgium)
Slideshare MyElaN in French (Belgium)Slideshare MyElaN in French (Belgium)
Slideshare MyElaN in French (Belgium)ElaN Languages
 
Week 1 discussion 2
Week 1  discussion 2Week 1  discussion 2
Week 1 discussion 2Shay89
 
Projet CogLab - PPT1
Projet CogLab - PPT1Projet CogLab - PPT1
Projet CogLab - PPT1
af83
 
"Internet revoluciona China" | Hispanohablantes en Asia > Comunidad Global > ...
"Internet revoluciona China" | Hispanohablantes en Asia > Comunidad Global > ..."Internet revoluciona China" | Hispanohablantes en Asia > Comunidad Global > ...
"Internet revoluciona China" | Hispanohablantes en Asia > Comunidad Global > ...
Roger Nierga
 
Manual cop-dvr16 rs-cop-dvr16hdmi
Manual cop-dvr16 rs-cop-dvr16hdmiManual cop-dvr16 rs-cop-dvr16hdmi
Manual cop-dvr16 rs-cop-dvr16hdmi
falames
 
Metodo del Camino Critico CPM PERT Arq. Derby Gonzalez INTEC
Metodo del Camino Critico CPM PERT Arq. Derby Gonzalez INTECMetodo del Camino Critico CPM PERT Arq. Derby Gonzalez INTEC
Metodo del Camino Critico CPM PERT Arq. Derby Gonzalez INTEC
Fred Pezoa
 
Open networking - BNI Insomniacs
Open networking - BNI InsomniacsOpen networking - BNI Insomniacs
Open networking - BNI Insomniacs
Muneer Samnani
 

Viewers also liked (20)

Curso auxiliar de enfermería
Curso auxiliar de enfermeríaCurso auxiliar de enfermería
Curso auxiliar de enfermería
 
ONG and crowdfunding case - 2012
ONG and crowdfunding case - 2012ONG and crowdfunding case - 2012
ONG and crowdfunding case - 2012
 
Asistencia de cabildo 06
Asistencia de cabildo 06Asistencia de cabildo 06
Asistencia de cabildo 06
 
Cv B Marco Crb Rev6 Ij
Cv B Marco Crb Rev6 IjCv B Marco Crb Rev6 Ij
Cv B Marco Crb Rev6 Ij
 
Encontro com a escritora Rosa Duarte
Encontro com a escritora Rosa Duarte Encontro com a escritora Rosa Duarte
Encontro com a escritora Rosa Duarte
 
Magazine con alma de blues n18
Magazine con alma de blues n18Magazine con alma de blues n18
Magazine con alma de blues n18
 
Red de rarea local2
Red de rarea local2Red de rarea local2
Red de rarea local2
 
EnerEscolas
EnerEscolasEnerEscolas
EnerEscolas
 
HBSL - Presentation for Educational Institutions March-2016
HBSL - Presentation for Educational Institutions March-2016HBSL - Presentation for Educational Institutions March-2016
HBSL - Presentation for Educational Institutions March-2016
 
Normas Insutec Virtual
Normas Insutec VirtualNormas Insutec Virtual
Normas Insutec Virtual
 
[En]ICFecc 2010 sponsor
[En]ICFecc 2010 sponsor[En]ICFecc 2010 sponsor
[En]ICFecc 2010 sponsor
 
Omer Kalil Testimony
Omer Kalil TestimonyOmer Kalil Testimony
Omer Kalil Testimony
 
Slideshare MyElaN in French (Belgium)
Slideshare MyElaN in French (Belgium)Slideshare MyElaN in French (Belgium)
Slideshare MyElaN in French (Belgium)
 
Week 1 discussion 2
Week 1  discussion 2Week 1  discussion 2
Week 1 discussion 2
 
Projet CogLab - PPT1
Projet CogLab - PPT1Projet CogLab - PPT1
Projet CogLab - PPT1
 
"Internet revoluciona China" | Hispanohablantes en Asia > Comunidad Global > ...
"Internet revoluciona China" | Hispanohablantes en Asia > Comunidad Global > ..."Internet revoluciona China" | Hispanohablantes en Asia > Comunidad Global > ...
"Internet revoluciona China" | Hispanohablantes en Asia > Comunidad Global > ...
 
Searl pk
Searl pkSearl pk
Searl pk
 
Manual cop-dvr16 rs-cop-dvr16hdmi
Manual cop-dvr16 rs-cop-dvr16hdmiManual cop-dvr16 rs-cop-dvr16hdmi
Manual cop-dvr16 rs-cop-dvr16hdmi
 
Metodo del Camino Critico CPM PERT Arq. Derby Gonzalez INTEC
Metodo del Camino Critico CPM PERT Arq. Derby Gonzalez INTECMetodo del Camino Critico CPM PERT Arq. Derby Gonzalez INTEC
Metodo del Camino Critico CPM PERT Arq. Derby Gonzalez INTEC
 
Open networking - BNI Insomniacs
Open networking - BNI InsomniacsOpen networking - BNI Insomniacs
Open networking - BNI Insomniacs
 

Similar to Exploring Strategies for Training Deep Neural Networks paper review

Fundamental of deep learning
Fundamental of deep learningFundamental of deep learning
Fundamental of deep learning
Stanley Wang
 
Autoencoders for image_classification
Autoencoders for image_classificationAutoencoders for image_classification
Autoencoders for image_classification
Cenk Bircanoğlu
 
DSRLab seminar Introduction to deep learning
DSRLab seminar   Introduction to deep learningDSRLab seminar   Introduction to deep learning
DSRLab seminar Introduction to deep learning
Poo Kuan Hoong
 
CNN.pptx.pdf
CNN.pptx.pdfCNN.pptx.pdf
CNN.pptx.pdf
Knoldus Inc.
 
Nips 2017 in a nutshell
Nips 2017 in a nutshellNips 2017 in a nutshell
Nips 2017 in a nutshell
LULU CHENG
 
deeplearning
deeplearningdeeplearning
deeplearning
huda2018
 
Practical ML
Practical MLPractical ML
Practical ML
Antonio Pitasi
 
Deep learning from a novice perspective
Deep learning from a novice perspectiveDeep learning from a novice perspective
Deep learning from a novice perspective
Anirban Santara
 
ML Module 3 Non Linear Learning.pptx
ML Module 3 Non Linear Learning.pptxML Module 3 Non Linear Learning.pptx
ML Module 3 Non Linear Learning.pptx
DebabrataPain1
 
Efficient design of feedforward network for pattern classification
Efficient design of feedforward network for pattern classificationEfficient design of feedforward network for pattern classification
Efficient design of feedforward network for pattern classification
IOSR Journals
 
GNR638_Course Project for spring semester
GNR638_Course Project for spring semesterGNR638_Course Project for spring semester
GNR638_Course Project for spring semester
BijayChandraDasTECH0
 
Objective Evaluation of a Deep Neural Network Approach for Single-Channel Spe...
Objective Evaluation of a Deep Neural Network Approach for Single-Channel Spe...Objective Evaluation of a Deep Neural Network Approach for Single-Channel Spe...
Objective Evaluation of a Deep Neural Network Approach for Single-Channel Spe...
csandit
 
V2.0 open power ai virtual university deep learning and ai introduction
V2.0 open power ai virtual university   deep learning and ai introductionV2.0 open power ai virtual university   deep learning and ai introduction
V2.0 open power ai virtual university deep learning and ai introduction
Ganesan Narayanasamy
 
Deep Learning via Semi-Supervised Embedding (第 7 回 Deep Learning 勉強会資料; 大澤)
Deep Learning via Semi-Supervised Embedding (第 7 回 Deep Learning 勉強会資料; 大澤)Deep Learning via Semi-Supervised Embedding (第 7 回 Deep Learning 勉強会資料; 大澤)
Deep Learning via Semi-Supervised Embedding (第 7 回 Deep Learning 勉強会資料; 大澤)
Ohsawa Goodfellow
 
DMS MODULE 1 PRESENTATION.pptx
DMS MODULE 1 PRESENTATION.pptxDMS MODULE 1 PRESENTATION.pptx
DMS MODULE 1 PRESENTATION.pptx
SREESAIARJUNKOSINEPA
 
GNR638_project ppt.pdf
GNR638_project ppt.pdfGNR638_project ppt.pdf
GNR638_project ppt.pdf
AtulVerma631398
 
PR-344: A Battle of Network Structures: An Empirical Study of CNN, Transforme...
PR-344: A Battle of Network Structures: An Empirical Study of CNN, Transforme...PR-344: A Battle of Network Structures: An Empirical Study of CNN, Transforme...
PR-344: A Battle of Network Structures: An Empirical Study of CNN, Transforme...
Jinwon Lee
 
2017 (albawi-alkabi)image-net classification with deep convolutional neural n...
2017 (albawi-alkabi)image-net classification with deep convolutional neural n...2017 (albawi-alkabi)image-net classification with deep convolutional neural n...
2017 (albawi-alkabi)image-net classification with deep convolutional neural n...
ali hassan
 
Deep Learning for NLP (without Magic) - Richard Socher and Christopher Manning
Deep Learning for NLP (without Magic) - Richard Socher and Christopher ManningDeep Learning for NLP (without Magic) - Richard Socher and Christopher Manning
Deep Learning for NLP (without Magic) - Richard Socher and Christopher Manning
BigDataCloud
 

Similar to Exploring Strategies for Training Deep Neural Networks paper review (20)

Fundamental of deep learning
Fundamental of deep learningFundamental of deep learning
Fundamental of deep learning
 
Autoencoders for image_classification
Autoencoders for image_classificationAutoencoders for image_classification
Autoencoders for image_classification
 
DSRLab seminar Introduction to deep learning
DSRLab seminar   Introduction to deep learningDSRLab seminar   Introduction to deep learning
DSRLab seminar Introduction to deep learning
 
CNN.pptx.pdf
CNN.pptx.pdfCNN.pptx.pdf
CNN.pptx.pdf
 
Nips 2017 in a nutshell
Nips 2017 in a nutshellNips 2017 in a nutshell
Nips 2017 in a nutshell
 
deeplearning
deeplearningdeeplearning
deeplearning
 
Practical ML
Practical MLPractical ML
Practical ML
 
Deep learning from a novice perspective
Deep learning from a novice perspectiveDeep learning from a novice perspective
Deep learning from a novice perspective
 
ML Module 3 Non Linear Learning.pptx
ML Module 3 Non Linear Learning.pptxML Module 3 Non Linear Learning.pptx
ML Module 3 Non Linear Learning.pptx
 
Efficient design of feedforward network for pattern classification
Efficient design of feedforward network for pattern classificationEfficient design of feedforward network for pattern classification
Efficient design of feedforward network for pattern classification
 
GNR638_Course Project for spring semester
GNR638_Course Project for spring semesterGNR638_Course Project for spring semester
GNR638_Course Project for spring semester
 
Objective Evaluation of a Deep Neural Network Approach for Single-Channel Spe...
Objective Evaluation of a Deep Neural Network Approach for Single-Channel Spe...Objective Evaluation of a Deep Neural Network Approach for Single-Channel Spe...
Objective Evaluation of a Deep Neural Network Approach for Single-Channel Spe...
 
V2.0 open power ai virtual university deep learning and ai introduction
V2.0 open power ai virtual university   deep learning and ai introductionV2.0 open power ai virtual university   deep learning and ai introduction
V2.0 open power ai virtual university deep learning and ai introduction
 
Deep Learning via Semi-Supervised Embedding (第 7 回 Deep Learning 勉強会資料; 大澤)
Deep Learning via Semi-Supervised Embedding (第 7 回 Deep Learning 勉強会資料; 大澤)Deep Learning via Semi-Supervised Embedding (第 7 回 Deep Learning 勉強会資料; 大澤)
Deep Learning via Semi-Supervised Embedding (第 7 回 Deep Learning 勉強会資料; 大澤)
 
DMS MODULE 1 PRESENTATION.pptx
DMS MODULE 1 PRESENTATION.pptxDMS MODULE 1 PRESENTATION.pptx
DMS MODULE 1 PRESENTATION.pptx
 
GNR638_project ppt.pdf
GNR638_project ppt.pdfGNR638_project ppt.pdf
GNR638_project ppt.pdf
 
PR-344: A Battle of Network Structures: An Empirical Study of CNN, Transforme...
PR-344: A Battle of Network Structures: An Empirical Study of CNN, Transforme...PR-344: A Battle of Network Structures: An Empirical Study of CNN, Transforme...
PR-344: A Battle of Network Structures: An Empirical Study of CNN, Transforme...
 
2017 (albawi-alkabi)image-net classification with deep convolutional neural n...
2017 (albawi-alkabi)image-net classification with deep convolutional neural n...2017 (albawi-alkabi)image-net classification with deep convolutional neural n...
2017 (albawi-alkabi)image-net classification with deep convolutional neural n...
 
Deep Learning for NLP (without Magic) - Richard Socher and Christopher Manning
Deep Learning for NLP (without Magic) - Richard Socher and Christopher ManningDeep Learning for NLP (without Magic) - Richard Socher and Christopher Manning
Deep Learning for NLP (without Magic) - Richard Socher and Christopher Manning
 
D028018022
D028018022D028018022
D028018022
 

More from Vimukthi Wickramasinghe

Beanstalkg
BeanstalkgBeanstalkg
Factored Operating Systems paper review
Factored Operating Systems paper reviewFactored Operating Systems paper review
Factored Operating Systems paper review
Vimukthi Wickramasinghe
 
Learning New Semi-Supervised Deep Auto-encoder Features for Statistical Machi...
Learning New Semi-Supervised Deep Auto-encoder Features for Statistical Machi...Learning New Semi-Supervised Deep Auto-encoder Features for Statistical Machi...
Learning New Semi-Supervised Deep Auto-encoder Features for Statistical Machi...
Vimukthi Wickramasinghe
 
Application Performance & Flexibility on Exokernel Systems paper review
Application Performance & Flexibility on Exokernel Systems paper reviewApplication Performance & Flexibility on Exokernel Systems paper review
Application Performance & Flexibility on Exokernel Systems paper review
Vimukthi Wickramasinghe
 
Improved Query Performance With Variant Indexes - review presentation
Improved Query Performance With Variant Indexes - review presentationImproved Query Performance With Variant Indexes - review presentation
Improved Query Performance With Variant Indexes - review presentation
Vimukthi Wickramasinghe
 
A parallel gpu version of the traveling salesman problem slides
A parallel gpu version of the traveling salesman problem slidesA parallel gpu version of the traveling salesman problem slides
A parallel gpu version of the traveling salesman problem slides
Vimukthi Wickramasinghe
 
Smart mrs bi project-presentation
Smart mrs bi project-presentationSmart mrs bi project-presentation
Smart mrs bi project-presentation
Vimukthi Wickramasinghe
 

More from Vimukthi Wickramasinghe (8)

Beanstalkg
BeanstalkgBeanstalkg
Beanstalkg
 
pgdip-project-report-final-148245F
pgdip-project-report-final-148245Fpgdip-project-report-final-148245F
pgdip-project-report-final-148245F
 
Factored Operating Systems paper review
Factored Operating Systems paper reviewFactored Operating Systems paper review
Factored Operating Systems paper review
 
Learning New Semi-Supervised Deep Auto-encoder Features for Statistical Machi...
Learning New Semi-Supervised Deep Auto-encoder Features for Statistical Machi...Learning New Semi-Supervised Deep Auto-encoder Features for Statistical Machi...
Learning New Semi-Supervised Deep Auto-encoder Features for Statistical Machi...
 
Application Performance & Flexibility on Exokernel Systems paper review
Application Performance & Flexibility on Exokernel Systems paper reviewApplication Performance & Flexibility on Exokernel Systems paper review
Application Performance & Flexibility on Exokernel Systems paper review
 
Improved Query Performance With Variant Indexes - review presentation
Improved Query Performance With Variant Indexes - review presentationImproved Query Performance With Variant Indexes - review presentation
Improved Query Performance With Variant Indexes - review presentation
 
A parallel gpu version of the traveling salesman problem slides
A parallel gpu version of the traveling salesman problem slidesA parallel gpu version of the traveling salesman problem slides
A parallel gpu version of the traveling salesman problem slides
 
Smart mrs bi project-presentation
Smart mrs bi project-presentationSmart mrs bi project-presentation
Smart mrs bi project-presentation
 

Recently uploaded

Basic Industrial Engineering terms for apparel
Basic Industrial Engineering terms for apparelBasic Industrial Engineering terms for apparel
Basic Industrial Engineering terms for apparel
top1002
 
space technology lecture notes on satellite
space technology lecture notes on satellitespace technology lecture notes on satellite
space technology lecture notes on satellite
ongomchris
 
ML for identifying fraud using open blockchain data.pptx
ML for identifying fraud using open blockchain data.pptxML for identifying fraud using open blockchain data.pptx
ML for identifying fraud using open blockchain data.pptx
Vijay Dialani, PhD
 
14 Template Contractual Notice - EOT Application
14 Template Contractual Notice - EOT Application14 Template Contractual Notice - EOT Application
14 Template Contractual Notice - EOT Application
SyedAbiiAzazi1
 
Hierarchical Digital Twin of a Naval Power System
Hierarchical Digital Twin of a Naval Power SystemHierarchical Digital Twin of a Naval Power System
Hierarchical Digital Twin of a Naval Power System
Kerry Sado
 
一比一原版(IIT毕业证)伊利诺伊理工大学毕业证成绩单专业办理
一比一原版(IIT毕业证)伊利诺伊理工大学毕业证成绩单专业办理一比一原版(IIT毕业证)伊利诺伊理工大学毕业证成绩单专业办理
一比一原版(IIT毕业证)伊利诺伊理工大学毕业证成绩单专业办理
zwunae
 
Water Industry Process Automation and Control Monthly - May 2024.pdf
Water Industry Process Automation and Control Monthly - May 2024.pdfWater Industry Process Automation and Control Monthly - May 2024.pdf
Water Industry Process Automation and Control Monthly - May 2024.pdf
Water Industry Process Automation & Control
 
Unbalanced Three Phase Systems and circuits.pptx
Unbalanced Three Phase Systems and circuits.pptxUnbalanced Three Phase Systems and circuits.pptx
Unbalanced Three Phase Systems and circuits.pptx
ChristineTorrepenida1
 
6th International Conference on Machine Learning & Applications (CMLA 2024)
6th International Conference on Machine Learning & Applications (CMLA 2024)6th International Conference on Machine Learning & Applications (CMLA 2024)
6th International Conference on Machine Learning & Applications (CMLA 2024)
ClaraZara1
 
Nuclear Power Economics and Structuring 2024
Nuclear Power Economics and Structuring 2024Nuclear Power Economics and Structuring 2024
Nuclear Power Economics and Structuring 2024
Massimo Talia
 
Student information management system project report ii.pdf
Student information management system project report ii.pdfStudent information management system project report ii.pdf
Student information management system project report ii.pdf
Kamal Acharya
 
weather web application report.pdf
weather web application report.pdfweather web application report.pdf
weather web application report.pdf
Pratik Pawar
 
Investor-Presentation-Q1FY2024 investor presentation document.pptx
Investor-Presentation-Q1FY2024 investor presentation document.pptxInvestor-Presentation-Q1FY2024 investor presentation document.pptx
Investor-Presentation-Q1FY2024 investor presentation document.pptx
AmarGB2
 
Immunizing Image Classifiers Against Localized Adversary Attacks
Immunizing Image Classifiers Against Localized Adversary AttacksImmunizing Image Classifiers Against Localized Adversary Attacks
Immunizing Image Classifiers Against Localized Adversary Attacks
gerogepatton
 
Planning Of Procurement o different goods and services
Planning Of Procurement o different goods and servicesPlanning Of Procurement o different goods and services
Planning Of Procurement o different goods and services
JoytuBarua2
 
Pile Foundation by Venkatesh Taduvai (Sub Geotechnical Engineering II)-conver...
Pile Foundation by Venkatesh Taduvai (Sub Geotechnical Engineering II)-conver...Pile Foundation by Venkatesh Taduvai (Sub Geotechnical Engineering II)-conver...
Pile Foundation by Venkatesh Taduvai (Sub Geotechnical Engineering II)-conver...
AJAYKUMARPUND1
 
Design and Analysis of Algorithms-DP,Backtracking,Graphs,B&B
Design and Analysis of Algorithms-DP,Backtracking,Graphs,B&BDesign and Analysis of Algorithms-DP,Backtracking,Graphs,B&B
Design and Analysis of Algorithms-DP,Backtracking,Graphs,B&B
Sreedhar Chowdam
 
Sachpazis:Terzaghi Bearing Capacity Estimation in simple terms with Calculati...
Sachpazis:Terzaghi Bearing Capacity Estimation in simple terms with Calculati...Sachpazis:Terzaghi Bearing Capacity Estimation in simple terms with Calculati...
Sachpazis:Terzaghi Bearing Capacity Estimation in simple terms with Calculati...
Dr.Costas Sachpazis
 
AKS UNIVERSITY Satna Final Year Project By OM Hardaha.pdf
AKS UNIVERSITY Satna Final Year Project By OM Hardaha.pdfAKS UNIVERSITY Satna Final Year Project By OM Hardaha.pdf
AKS UNIVERSITY Satna Final Year Project By OM Hardaha.pdf
SamSarthak3
 
在线办理(ANU毕业证书)澳洲国立大学毕业证录取通知书一模一样
在线办理(ANU毕业证书)澳洲国立大学毕业证录取通知书一模一样在线办理(ANU毕业证书)澳洲国立大学毕业证录取通知书一模一样
在线办理(ANU毕业证书)澳洲国立大学毕业证录取通知书一模一样
obonagu
 

Recently uploaded (20)

Basic Industrial Engineering terms for apparel
Basic Industrial Engineering terms for apparelBasic Industrial Engineering terms for apparel
Basic Industrial Engineering terms for apparel
 
space technology lecture notes on satellite
space technology lecture notes on satellitespace technology lecture notes on satellite
space technology lecture notes on satellite
 
ML for identifying fraud using open blockchain data.pptx
ML for identifying fraud using open blockchain data.pptxML for identifying fraud using open blockchain data.pptx
ML for identifying fraud using open blockchain data.pptx
 
14 Template Contractual Notice - EOT Application
14 Template Contractual Notice - EOT Application14 Template Contractual Notice - EOT Application
14 Template Contractual Notice - EOT Application
 
Hierarchical Digital Twin of a Naval Power System
Hierarchical Digital Twin of a Naval Power SystemHierarchical Digital Twin of a Naval Power System
Hierarchical Digital Twin of a Naval Power System
 
一比一原版(IIT毕业证)伊利诺伊理工大学毕业证成绩单专业办理
一比一原版(IIT毕业证)伊利诺伊理工大学毕业证成绩单专业办理一比一原版(IIT毕业证)伊利诺伊理工大学毕业证成绩单专业办理
一比一原版(IIT毕业证)伊利诺伊理工大学毕业证成绩单专业办理
 
Water Industry Process Automation and Control Monthly - May 2024.pdf
Water Industry Process Automation and Control Monthly - May 2024.pdfWater Industry Process Automation and Control Monthly - May 2024.pdf
Water Industry Process Automation and Control Monthly - May 2024.pdf
 
Unbalanced Three Phase Systems and circuits.pptx
Unbalanced Three Phase Systems and circuits.pptxUnbalanced Three Phase Systems and circuits.pptx
Unbalanced Three Phase Systems and circuits.pptx
 
6th International Conference on Machine Learning & Applications (CMLA 2024)
6th International Conference on Machine Learning & Applications (CMLA 2024)6th International Conference on Machine Learning & Applications (CMLA 2024)
6th International Conference on Machine Learning & Applications (CMLA 2024)
 
Nuclear Power Economics and Structuring 2024
Nuclear Power Economics and Structuring 2024Nuclear Power Economics and Structuring 2024
Nuclear Power Economics and Structuring 2024
 
Student information management system project report ii.pdf
Student information management system project report ii.pdfStudent information management system project report ii.pdf
Student information management system project report ii.pdf
 
weather web application report.pdf
weather web application report.pdfweather web application report.pdf
weather web application report.pdf
 
Investor-Presentation-Q1FY2024 investor presentation document.pptx
Investor-Presentation-Q1FY2024 investor presentation document.pptxInvestor-Presentation-Q1FY2024 investor presentation document.pptx
Investor-Presentation-Q1FY2024 investor presentation document.pptx
 
Immunizing Image Classifiers Against Localized Adversary Attacks
Immunizing Image Classifiers Against Localized Adversary AttacksImmunizing Image Classifiers Against Localized Adversary Attacks
Immunizing Image Classifiers Against Localized Adversary Attacks
 
Planning Of Procurement o different goods and services
Planning Of Procurement o different goods and servicesPlanning Of Procurement o different goods and services
Planning Of Procurement o different goods and services
 
Pile Foundation by Venkatesh Taduvai (Sub Geotechnical Engineering II)-conver...
Pile Foundation by Venkatesh Taduvai (Sub Geotechnical Engineering II)-conver...Pile Foundation by Venkatesh Taduvai (Sub Geotechnical Engineering II)-conver...
Pile Foundation by Venkatesh Taduvai (Sub Geotechnical Engineering II)-conver...
 
Design and Analysis of Algorithms-DP,Backtracking,Graphs,B&B
Design and Analysis of Algorithms-DP,Backtracking,Graphs,B&BDesign and Analysis of Algorithms-DP,Backtracking,Graphs,B&B
Design and Analysis of Algorithms-DP,Backtracking,Graphs,B&B
 
Sachpazis:Terzaghi Bearing Capacity Estimation in simple terms with Calculati...
Sachpazis:Terzaghi Bearing Capacity Estimation in simple terms with Calculati...Sachpazis:Terzaghi Bearing Capacity Estimation in simple terms with Calculati...
Sachpazis:Terzaghi Bearing Capacity Estimation in simple terms with Calculati...
 
AKS UNIVERSITY Satna Final Year Project By OM Hardaha.pdf
AKS UNIVERSITY Satna Final Year Project By OM Hardaha.pdfAKS UNIVERSITY Satna Final Year Project By OM Hardaha.pdf
AKS UNIVERSITY Satna Final Year Project By OM Hardaha.pdf
 
在线办理(ANU毕业证书)澳洲国立大学毕业证录取通知书一模一样
在线办理(ANU毕业证书)澳洲国立大学毕业证录取通知书一模一样在线办理(ANU毕业证书)澳洲国立大学毕业证录取通知书一模一样
在线办理(ANU毕业证书)澳洲国立大学毕业证录取通知书一模一样
 

Exploring Strategies for Training Deep Neural Networks paper review

  • 1. Exploring Strategies for Training Deep Neural Networks By Hugo Larochelle, Yoshua Bengio,Jerome Louradour, Pascal Lamblin By V B Wickramasinghe (148245F)
  • 2. Outline ● Introduction ● Deep neural networks ● Stacked Restricted Boltzmann Machine Network ● Stacked Autoassociators Network ● Experimental results ● Conclusion
  • 3. Introduction ● Training deep neural network is hard. ● This is mainly due to randomly initialized deep architecture tend to get stuck in poor situations. ● But the ability of deep architectures to represent complex functions is unmatched. ● This paper highlights some of the recent breakthroughs in training deep architectures that has helped to uncover their potential.
  • 4. Deep neural networks ● Shallow networks has been proved to be inefficient in circuit theory, boolean logic and neural networks. ● This is because some functions that can be represented using k layers is with finite number of units takes exponential number units with k-1 layers. ● Also highly varying function can be easily represented by a number of non-linearities stacked together. ● Another issue with shallow architectures is that they’ll require exponential number of training examples to learn complex functions ● But as mentioned earlier training deep architectures is hard. What is the solution?
  • 6. Stacked Restricted Boltzmann Machine Network ● RBMs represent a generative model of input. ● Train individual layers of RBMs using contrastive divergence. ● Then stack them together so that a one layers output representation works as input to another(A DBN). ● Hinton(2006) argues that this helps in a more complex representation overall. ● Then the pretrained stacked framework can be trained to for a particular task using backpropagation.
  • 7. Stacked Autoassociators Network ● Like RBMs autoassociators are a type of network that when combined helps improving input representation. ● Autoassociators are an encoding model which is trained to minimize the reconstruction loss of input from output. ● Stacked autoassociator performs same layer wise training procedure as DBNs. ● Reconstruction error of an autoassociator and log-likelihood of RBM are both approximate values of convergent series of log-likelihood gradient obtained in different ways.
  • 13. Conclusion ● DNNs are an indispensable tool for learning tasks. ● This paper presents 3 methods of optimally training DNNs, 1. pre-training one layer at a time in a greedy way. 2. using unsupervised learning at each layer in a way that preserves information from the input and disentangles factors of variation. 3. fine-tuning the whole network with respect to the ultimate criterion of interest. ● The experiments are sound and present clearly why deep neural networks trained using the presented methods can help in improving learning tasks significantly over single layer networks.