Presentation is loading. Please wait.

Presentation is loading. Please wait.

Comp 3503 / 5013 Dynamic Neural Networks Daniel L. Silver March, 2014 1.

Similar presentations


Presentation on theme: "Comp 3503 / 5013 Dynamic Neural Networks Daniel L. Silver March, 2014 1."— Presentation transcript:

1 Comp 3503 / 5013 Dynamic Neural Networks Daniel L. Silver March, 2014 1

2 Outline Hopfield Networks Boltzman Machines Mean Field Theory Restricted Boltzman Machines (RBM) 2

3 Dynamic Neural Networks See handout for image of spider, beer and dog The search for a model or hypothesis can be considered the relaxation of a dynamic system into a state of equilibrium This is the nature of most physical systems – Pool of water – Air in a room Mathematics is that of thermal-dynamics – Quote from John Von Neumann 3

4 Hopfield Networks See hand out 4

5 Hopfield Networks Hopfield Network video intro – http://www.youtube.com/watch?v=gfPUWwBkXZ Y http://www.youtube.com/watch?v=gfPUWwBkXZ Y – http://faculty.etsu.edu/knisleyj/neural/ http://faculty.etsu.edu/knisleyj/neural/ Try these Applets: – http://lcn.epfl.ch/tutorial/english/hopfield/html/i ndex.html http://lcn.epfl.ch/tutorial/english/hopfield/html/i ndex.html – http://www.cbu.edu/~pong/ai/hopfield/hopfielda pplet.html http://www.cbu.edu/~pong/ai/hopfield/hopfielda pplet.html 5

6 Hopfield Networks Basics with Geoff Hinton: Introduction to Hopfield Nets – http://www.youtube.com/watch?v=YB3-Hn-inHI http://www.youtube.com/watch?v=YB3-Hn-inHI Storage capacity of Hopfield Nets – http://www.youtube.com/watch?v=O1rPQlKQBLQ http://www.youtube.com/watch?v=O1rPQlKQBLQ 6

7 Hopfield Networks Advanced concepts with Geoff Hinton: Hopfield nets with hidden units – http://www.youtube.com/watch?v=bOpddsa4BPI http://www.youtube.com/watch?v=bOpddsa4BPI Necker Cube – http://www.cs.cf.ac.uk/Dave/JAVA/boltzman/Neck er.html http://www.cs.cf.ac.uk/Dave/JAVA/boltzman/Neck er.html Adding noise to improve search – http://www.youtube.com/watch?v=kVgT2Eaa6KA http://www.youtube.com/watch?v=kVgT2Eaa6KA 7

8 Boltzman Machine -See Handout -http://www.scholarpedia.org/article/Boltzmann_machinehttp://www.scholarpedia.org/article/Boltzmann_machine Basics with Geoff Hinton Modeling binary data – http://www.youtube.com/watch?v=MKdvJst8a6k http://www.youtube.com/watch?v=MKdvJst8a6k BM Learning Algorithm – http://www.youtube.com/watch?v=QgrFsnHFeig http://www.youtube.com/watch?v=QgrFsnHFeig 8

9 Limitations of BMs BM Learning does not scale well This is due to several factors, the most important being: – The time the machine must be run in order to collect equilibrium statistics grows exponentially with the machine's size = number of nodes For each example – sample nodes, sample states – Connection strengths are more plastic when the units have activation probabilities intermediate between zero and one. Noise causes the weights to follow a random walk until the activities saturate (variance trap). 9

10 Potential Solutions Use a momentum term as in BP: Add a penalty term to create sparse coding (encourage shorter encodings for different inputs) Use implementation tricks to do more in memory – batches of examples Restrict number of iterations in + and – phases Restrict connectivity of network 10 w ij (t+1)=w ij (t) +ηΔw ij +αΔw ij (t-1)

11 Restricted Boltzman Machine 11 Source: http://blog.echen.me/2011/07/18/introduction-to-restricted-boltzmann-machines/http://blog.echen.me/2011/07/18/introduction-to-restricted-boltzmann-machines/ SF/Fantasy Oscar Winner w ij j i Σ j =w ij v i h j  p j =1/(1-e -Σj ) v i  p i =1/(1-e -Σi ) Recall = Relaxation Σ i =w ij h j v o or h o

12 Restricted Boltzman Machine 12 Source: http://blog.echen.me/2011/07/18/introduction-to-restricted-boltzmann-machines/http://blog.echen.me/2011/07/18/introduction-to-restricted-boltzmann-machines/ SF/Fantasy Oscar Winner w ij j i Σ j =w ij v i h j  p j =1/(1-e -Σj ) v i  p i =1/(1-e -Σi ) Recall = Relaxation Σ i =w ij h j v o or h o

13 Restricted Boltzman Machine 13 Source: http://blog.echen.me/2011/07/18/introduction-to-restricted-boltzmann-machines/http://blog.echen.me/2011/07/18/introduction-to-restricted-boltzmann-machines/ SF/Fantasy Oscar Winner j i h j  p j =1/(1-e -Σj ) v i  p i =1/(1-e -Σi ) Σ i =w ij h j v o or h o Oscar Winner SF/Fantasy Recall = Relaxation w ij Σ j =w ij v i

14 Restricted Boltzman Machine 14 Source: http://blog.echen.me/2011/07/18/introduction-to-restricted-boltzmann-machines/http://blog.echen.me/2011/07/18/introduction-to-restricted-boltzmann-machines/ SF/Fantasy Oscar Winner j i h j  p j =1/(1-e -Σj ) v i  p i =1/(1-e -Σi ) Σ i =w ij h j v o or h o Oscar Winner SF/Fantasy Recall = Relaxation w ij Σ j =w ij v i

15 Restricted Boltzman Machine 15 Source: http://blog.echen.me/2011/07/18/introduction-to-restricted-boltzmann-machines/http://blog.echen.me/2011/07/18/introduction-to-restricted-boltzmann-machines/ SF/Fantasy Oscar Winner j i Σ i =w ij h j h j  p j =1/(1-e -Σj ) v i  p i =1/(1-e -Σi ) Learning = ~ Gradient Descent = Constrastive Divergence Update hidden units P=P+v i h j v o or h o Σ j =w ij v i

16 Restricted Boltzman Machine 16 Source: http://blog.echen.me/2011/07/18/introduction-to-restricted-boltzmann-machines/http://blog.echen.me/2011/07/18/introduction-to-restricted-boltzmann-machines/ SF/Fantasy Oscar Winner j i h j  p j =1/(1-e -Σj ) v i  p i =1/(1-e -Σi ) Learning = ~ Gradient Descent = Constrastive Divergence Reconstruct visible units v o or h o Σ j =w ij v i Σ i =w ij h j

17 Restricted Boltzman Machine 17 Source: http://blog.echen.me/2011/07/18/introduction-to-restricted-boltzmann-machines/http://blog.echen.me/2011/07/18/introduction-to-restricted-boltzmann-machines/ SF/Fantasy Oscar Winner j i Σ j =w ij v i h j  p j =1/(1-e -Σj ) v i  p i =1/(1-e -Σi ) Learning = ~ Gradient Descent = Constrastive Divergence Reupdate hidden units v o or h o Σ i =w ij h j N=N+v i h j

18 Restricted Boltzman Machine 18 Source: http://blog.echen.me/2011/07/18/introduction-to-restricted-boltzmann-machines/http://blog.echen.me/2011/07/18/introduction-to-restricted-boltzmann-machines/ SF/Fantasy Oscar Winner Δw ij = - j i Σ j =w ij v i h j  p j =1/(1-e -Σj ) v i  p i =1/(1-e -Σi ) Σ i =w ij h j v o or h o w ij =w ij +ηΔw ij Learning = ~ Gradient Descent = Constrastive Divergence Update weights

19 Restricted Boltzman Machine RBM Overview: – http://blog.echen.me/2011/07/18/introduction- to-restricted-boltzmann-machines/ http://blog.echen.me/2011/07/18/introduction- to-restricted-boltzmann-machines/ Wikipedia on DLA and RBM: – http://en.wikipedia.org/wiki/Deep_learning http://en.wikipedia.org/wiki/Deep_learning RBM Details and Code: – http://www.deeplearning.net/tutorial/rbm.html http://www.deeplearning.net/tutorial/rbm.html 19

20 Restricted Boltzman Machine Geoff Hinton on RBMs: RBMs and Constrastive Divergence Algorithm – http://www.youtube.com/watch?v=fJjkHAuW0Yk http://www.youtube.com/watch?v=fJjkHAuW0Yk An example of RBM Learning – http://www.youtube.com/watch?v=Ivj7jymShN0 http://www.youtube.com/watch?v=Ivj7jymShN0 RBMs applied to Collaborative Filtering – http://www.youtube.com/watch?v=laVC6WFIXjg http://www.youtube.com/watch?v=laVC6WFIXjg 20

21 Additional References Coursera course – Neural Networks fro Machine Learning: – https://class.coursera.org/neuralnets-2012- 001/lecture https://class.coursera.org/neuralnets-2012- 001/lecture ML: Hottest Tech Trend in next 3-5 Years – http://www.youtube.com/watch?v=b4zr9Zx5WiE http://www.youtube.com/watch?v=b4zr9Zx5WiE 21


Download ppt "Comp 3503 / 5013 Dynamic Neural Networks Daniel L. Silver March, 2014 1."

Similar presentations


Ads by Google