1 / 22

Outline

Online Dynamic Value System for Machine Learning Haibo He, Stevens Institute of Technology Janusz A. Starzyk, Ohio University. Outline. Introduction; Online curve fitting principles ; Network architecture and operation; Simulation analysis; Conclusion and future research;.

elpida
Download Presentation

Outline

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. Online Dynamic Value System for Machine LearningHaibo He, Stevens Institute of Technology Janusz A. Starzyk, Ohio University

  2. Outline • Introduction; • Online curve fitting principles; • Network architecture and operation; • Simulation analysis; • Conclusion and future research;

  3. Introduction: Why value system is important? • From traditional AI to the embodied intelligence: Rat Neurons can fly F- 22 jet Picture source:www.space.com • Make value judgments according to received information; • Develop sensory-motor coordination to actively interaction with environment; • Develop internal value system and apply it to decision making;

  4. Introduction: What is the value signal? • Different applications will have different definition of value signal, but we define the value signal as an expected reward or desired objective for machine’s action. Motivation: Goal-driven learning To provide a mechanism for the intelligent machines to be able to dynamically estimate the value function in reinforcement learning (specify “good” from “bad”), therefore guiding the machines to adjust its actions to achieve the goal. Source: Biologically inspired robot at CWRU http://biorobots.cwru.edu/

  5. Characteristics: * Self-organization * Sparse and local interconnections * Dynamically reconfigurable * Online data-driven learning Remote neurons System clock Nearest neighbour neuron Other Neurons II: information index ID: information deficiency Introduction: self-organizing learning array(SOLAR)

  6. How can value system help here? • Supervisor is not always available in the learning environment • Uncertain (no prior knowledge) external environment • Supervisor is not always necessary in the learning environment • How learning happens in a one-year old baby Source: Sociable humanoid robots: Kismet at MIT Artificial Intelligence Lab

  7. The challenges • Unstructured environment/uncertain information • Limited availability of information; • Information ambiguity and redundancy; • High dimensionality of the data set; • Time variability of the information;

  8. Outline • Introduction; • Online Curve Fitting Principles; • Network architecture and operation; • Simulation analysis; • Conclusion and future research;

  9. Online dynamic curve fitting Consider dynamic adjustment of the fit function described by a linear combination of the selected base functions: Storage requirements:

  10. Three curve fitting versus single curve fitting • Three curve fitting: • Neutral Curve: a least square fit (LSF) fits to all the data samples in the space • Upper Curve: only fits to the data points which are above the neutral curve. • Lower Curve: only fits to the data points which are below the neutral curve

  11. Decision integration Differential Based Voting:

  12. Implementation of TCF {New data sample comes; Modify the neutral curve; Difference = If (Difference >= 0) { Modify the lower curve; Keep the upper curve unchanged;} else { Modify the upper curve; Keep the lower curve unchanged;} end end} Pseudo code:

  13. Outline • Introduction; • Online Curve Fitting Principles; • Network architecture and operation; • Simulation analysis; • Conclusion and future research;

  14. Value system architecture A pipelined dynamic architecture:

  15. Inside a value system

  16. Outline • Introduction; • Online Curve Fitting Principles; • Network architecture and operation; • Simulation analysis; • Conclusion and future research;

  17. Simulation analysis Financial data analysis - bank prime loan rate prediction Data sets are available from: www.forecasts.org • Input: • Monthly bank prime loan rate; • Discount rate; • Federal funds rate; • Ten-year treasury constant maturity rate; • Output: • Next month’s bank prime loan rate • Training period: • January 1995 to December 2000 • Testing period: • February 2001 to September 2002 “market is unpredictable” • Random Walk Hypothesis; • Efficient Market Hypothesis;

  18. Prediction results Bank prime loan rate prediction by value system (February 2001 to September 2002)

  19. Result comparison: MSE error

  20. Outline • Introduction; • Online Curve Fitting Principles; • Network architecture and operation; • Simulation analysis; • Conclusion and future research;

  21. Conclusion and future research • Provide a mechanism for the intelligent machines to be able to dynamically estimate the value function; • Dynamic online data driven learning; • No backpropagation required; • Three curve fitting method; • General framework for different implementations

  22. Future research • Dynamically self-reconfigurable; • Investigate different input transformation and base functions; • Hardware implementation; • Facilitate goal-driven learning; • Integration with reinforcement learning within a realistic environment; A promising future? Ray Kurzweil predicted:We achieve one Human Brain capability for $1,000 around the year 2023, for one cent around the year 2037; We achieve one Human Race capability for $1,000 around the year 2049, for one cent around the year 2059. ---from “The Law of Accelerating Returns” by Ray Kurzweil Source: www.kurzweilai.net

More Related