1 / 15

A new type of Structured Artificial Neural Networks based on the Matrix Model of Computation

A new type of Structured Artificial Neural Networks based on the Matrix Model of Computation. Sergio Pissanetzky. The Matrix Model of Computation (MMC) consists of two sparse matrices: M = (C, Q) C = Matrix of Services Q = Matrix of Sequences.

bowen
Download Presentation

A new type of Structured Artificial Neural Networks based on the Matrix Model of Computation

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. A new type of Structured Artificial Neural Networksbased on the Matrix Model of Computation Sergio Pissanetzky

  2. The Matrix Model of Computation (MMC) consists of two sparse matrices: M = (C, Q) C = Matrix of Services Q = Matrix of Sequences

  3. The MMC is simple, yet very rich in features

  4. Converting a neural network to MMC b1 b2 w1 x1 w3 Σ φ1 v1 x3 w2 x2 φ2 Σ x5 v2 w4 x4

  5. MMC as new type of neural network ● MMC is equivalent to the equations● MMC supports all ANN features. ● In addition, MMC supports global features such as network structure and organization, objects and ontologies.

  6. The Scope Constriction Algorithm (SCA) SCA organizes information into objects and reveals the natural ontology of the system.

  7. PROGRAM DATA

  8. Profile

  9. Data channel “Turbulent” flow

  10. Service Commutativity

  11. PROGRAM DATA G “Laminar” flow H G H G H

  12. Where do objects come from? ● SCA only minimizes the scopes. ● In nature, scopes are resources. ● Processes compete for resources. ● Scopes are naturally minimized. ● Objects and inheritance arise naturally.

  13. DYNAMICS ● New information keeps arriving. ● SCA keeps forming new objects. ● Larger objects grow out of smaller objects. ● Objects evolve with time. Some stabilize. ● The process continues indefinitely.

  14. Lyapunov Dynamics vs. MMC Dynamics

  15. CONCLUSIONS AND OUTLOOK

More Related