Bob Wray, Randy Jones 8 Jun 2017

Slides:



Advertisements
Similar presentations
Training Guide. `
Advertisements

DARPA Mobile Autonomous Robot SoftwareMay Adaptive Intelligent Mobile Robotics William D. Smart, Presenter Leslie Pack Kaelbling, PI Artificial.
Chapter Five Evaluating your Skill- related Fitness.
4/15/2017 Using Gaussian Process Regression for Efficient Motion Planning in Environments with Deformable Objects Barbara Frank, Cyrill Stachniss, Nichola.
Computational Biology, Part 17 Biochemical Kinetics I Robert F. Murphy Copyright  1996, All rights reserved.
User Centered Design Lecture # 5 Gabriel Spitz.
U.S. Army Research, Development and Engineering Command Unclassified – Unlimited Distribution Considerations for adaptive tutoring within serious games:
Slide 1 Flight Simulation – Ch 8 A Case Study in an Architecture for “Integrability” Boeing 747 flight simulator in action – from
Instructional Design Eman Almasruhi.
Internet Based Information Sources on Urbanism - Tutorial - Authors: D. Milovanovic, D. S. Furundzic, yubc.net.
Sérgio Ronaldo Barros dos Santos (ITA-Brazil) Sidney Nascimento Givigi Júnior (RMC-Canada) Cairo Lúcio Nascimento Júnior (ITA-Brazil) Autonomous Construction.
Integrating Background Knowledge and Reinforcement Learning for Action Selection John E. Laird Nate Derbinsky Miller Tinkerhess.
1 S ystems Analysis Laboratory Helsinki University of Technology Kai Virtanen, Tuomas Raivio and Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki.
OntoSoar: Soar Finds Facts in Text Peter Lindes, Deryle Lonsdale, David Embley Brigham Young University 33 rd Soar Workshop, June 2013 pl 6/6/201333rd.
Copyright © 2010 by John Wiley & Sons, Inc. All Rights Reserved.
Commercially Available Patient Education Materials Module 11.
An instructional design theory for interactions in web-based learning environments 指導教授 : 陳 明 溥 研 究 生 : 許 良 村 Lee, M.& Paulus, T. (2001). An instructional.
Integration of generic competencies and content in the Skills for a Changing World Curriculum Key Issues to Consider.
Rob Byrd Chief Enterprise Architect Enterprise Architecture – A Citywide Service Delivery Strategy Aligning Information Technology Services to the Citizen.
Managerial Decision Making
Learning theory and ‘Career’
TRANSVERSAL COMPETENCIES IN EDUCATION POLICIES AND PRACTICE
Bob Wray, Charles Newton, Victor Hung, Norb Timpko 7 Jun 2017
Educational Psychology: Theory and Practice
User Documentation Stored information about how to use a system
Introduction to Employee Training and Development Chapter 1
The process of interaction design Chapter
DESIGNING IWB RESOURCES FOR LANGUAGE TEACHING: THE iTILT PROJECT
Chapter 5 – Requirements Engineering
Situated learning and Technology
LU4 Promoting Learning & Continuous Development Opportunities
Do it now activity Last lesson we used Flowol to create a solution to a problem a computer could solve. Identify what each symbol does:
Antecedent Control: Modeling, Guidance, and Situational Inducement
J.S. Bresee Conor McKenna iFest – 08/01/17
Implementing the NHS KSF Action Planning and Surgery Session
Building a Strong Relationship with the Business
The importance of emotional learning within communication between the staff Project Number: RO01-KA
PETRA 2014 An Interactive Learning and Adaptation Framework for Socially Assistive Robotics: An Interactive Reinforcement Learning Approach Konstantinos.
DEFECT PREDICTION : USING MACHINE LEARNING
Until Learning Has No Limits ™CAST
Overview of System Engineering
Best professional Pilot Training in India Airwing Aviation Academy Address: First Floor, Rachit Complex, Near Guru Nanak Girls.
TLO Action: Employ the Army Ethic in Leadership Positions
Strengths and Weaknesses
Teaching with Instructional Software
In this class…… We all have different starting points
A BRIEF INTRODUCTION TO UNIX OPERATING SYSTEM
Cognitive Language Comprehension in Rosie
MBML_Efficient Testing Methodology for Machine Learning
Soar Technology, Inc. Proprietary 11/26/2018
Soar Agents for Cyberspace 5/15/2018
Antecedent Control: Modeling, Guidance, and Situational Inducement
More Requirements Models
Machine Learning for Actuaries
Social Media Training Plan
STATE UNIVERSITY OF MAKASSAR 2011
Principles of Development
Warm Up Chapter 6.2 Separation of Variables Friday, April 12, 2019
Trust and Selling Ethically
ROLE OF «electronic virtual enhanced research-engaged student teams» WEB PORTAL IN SOLUTION OF PROBLEM OF COLLABORATION INTERNATIONAL TEAMS INSIDE ONE.
Introduction to Employee Training and Development Chapter 1
POWER CHALLENGES Several Ways To Solve 7 CHALLENGES.
New Challenges in Systems Safety - Themes
NAEYC Developmentally Appropriate Practice in Early Childhood Programs Serving Children from Birth through Age 8.
UPTIME & SEMANTIC WEB STANDARDS
Warm Up Siete Chapter 5.7 Separation of Variables Monday, May 27, 2019
Key Conclusions of Workshop Session 5: Research and future work needs – topics and ways of work.
Software Onboarding and Instructional Design
HUMAN COMPUTER INTERACTION. The main aims of the chapter are to: Explain the difference between good and poor interaction design. Describe what interaction.
Keys to Success in Engineering Study
Presentation transcript:

Bob Wray, Randy Jones 8 Jun 2017 Learning to “Soar” Creating families of models to support training Bob Wray, Randy Jones 8 Jun 2017 Copyright © 2017 Soar Technology, Inc.

Questions What are good (fast, cheap, reliable/robust) methods of developing/learning low-level robotic controller in Soar? What are good (fast, cheap, effective) methods for developing a family of Soar models?

Illustrative Problem Domain

Application Requirements Goal: Train an individual (“instructor”) who needs to interact with pilots with various skill levels Key aspects of training: Recognizing learner mistakes (e.g., control over-compensation) Learning when (and when not) to guide the learner explicitly Recognizing if the learner’s reactions to guidance are effective Requirements: Realistic flight (stick level control of aircraft) Generation of many different kinds of learner behaviors Interactive (responsive to instructor guidance)

Initial System Architecture

Question 1 What are good (fast, cheap, reliable/robust) methods of developing/learning low-level robotic controller in Soar? Generic joystick controller, rudder (slider) Obvious solution: Learn a “perfect” solution with RL? Computable, optimal flight paths enable straightforward formulation of policy for RL Examples of others who have used RL to learn low-level controllers? Are there other established ways to approach learning a robotic controller in Soar?

Question 2 What are good (fast, cheap, effective) methods for developing a family of Soar models? Does a traditional goal formulation + the RL policy result in learning that looks human like? Do flight paths generated by the system during RL look similar to human pilots along a similar learning path? Does variability in pilot performance arise from alternative goal decompositions? Are there ways to (easily) formulate “policies” for non-optimal flight paths?

System Architecture Questions

Conclusions Questions Nuggets Coal What are good (fast, cheap, reliable/robust) methods of developing/learning low-level robotic controller in Soar? What are good (fast, cheap, effective) methods for developing a family of Soar models? Nuggets Somewhat rare opportunity to build a Soar model of humans performing a complex perceptual/motor/decision task Opportunity to explore the trajectory of learning Coal Just getting started. Research, technology, and integration challenges ahead.