1 Chunking with Confidence John Laird University of Michigan June 17, 2005 25 th Soar Workshop

Slides:

Advertisements

Similar presentations

Time Management Tips for the Multitasking Research Administrator Source material taken from Eat that Frog! A book by Brian Tracy Presentation by Trish.

Advertisements

Heuristic Search techniques

Learning To Use Memory Nick Gorski & John Laird Soar Workshop 2011.

1 Update on Learning By Observation Learning from Positive Examples Only Tolga Konik University of Michigan.

RIPPER Fast Effective Rule Induction

Playing Dice with Soar Soar Workshop June 2011 John E. Laird Nate Derbinsky, Miller Tinkerhess University of Michigan 1.

Introduction to SOAR Based on “a gentle introduction to soar: an Architecture for Human Cognition” by Jill Fain Lehman, John Laird, Paul Rosenbloom. Presented.

SolidWorks Simulation. Dassault Systemes 3 – D and PLM software PLM - Product Lifecycle Management Building models on Computer Engineering Analysis and.

Evolutionary Computational Intelligence Lecture 10a: Surrogate Assisted Ferrante Neri University of Jyväskylä.

1 Episodic Memory for Soar Agents Andrew Nuxoll 11 June 2004.

Impact of Working Memory Activation on Agent Design John Laird, University of Michigan 1.

1 Soar Semantic Memory Yongjia Wang University of Michigan.

Redux Update: Building Rules from Examples Douglas Pearson John Laird ThreePenny Software University of Michigan

The Importance of Architecture for Achieving Human-level AI John Laird University of Michigan June 17, th Soar Workshop

A Soar’s Eye View of ACT-R John Laird 24 th Soar Workshop June 11, 2004.

Models of Human Performance Dr. Chris Baber. 2 Objectives Introduce theory-based models for predicting human performance Introduce competence-based models.

2004 Soar Technology, Inc.  10 Jun 2004  Robert Wray  Slide 1 A Goal Dependency Set Primer Robert Wray Soar Jun 2004.

D Nagesh Kumar, IIScOptimization Methods: M1L4 1 Introduction and Basic Concepts Classical and Advanced Techniques for Optimization.

Reinforcement Learning and Soar Shelley Nason. Reinforcement Learning Reinforcement learning: Learning how to act so as to maximize the expected cumulative.

Soar-RL: Reinforcement Learning and Soar Shelley Nason.

Solve an equation with variables on both sides

Order of Operations Arizona State 8 th Grade Math Standard – Strand 1, Concept 2, PO5 Simplify numerical expressions using the order of operations that.

COMPUTATIONAL MODELING OF INTEGRATED COGNITION AND EMOTION Bob MarinierUniversity of Michigan.

L-3 PROPRIETARY 1 Real-time IO-oriented Soar Agent for Base Station-Mobile Control David Daniel, Scott Hastings L-3 Communications.

Component Reliability Analysis

Knowledge Representation and Reasoning University "Politehnica" of Bucharest Department of Computer Science Fall 2010 Adina Magda Florea

THIS IS HOW WE DO IT! Logic and Information van Benthem-style Celebration Event in Honor of Johan van Benthem Amsterdam, September 26-27, 2014 Vincent.

Using the Order of Operations Mathematicians have established an order of operations to evaluate an expression involving more than one operation. Finally,

Using the Order of Operations Mathematicians have established an order of operations to evaluate an expression involving more than one operation. Finally,

GATree: Genetically Evolved Decision Trees 전자전기컴퓨터공학과 데이터베이스 연구실 G 김태종.

Integrating Background Knowledge and Reinforcement Learning for Action Selection John E. Laird Nate Derbinsky Miller Tinkerhess.

Decision Trees DefinitionDefinition MechanismMechanism Splitting FunctionSplitting Function Issues in Decision-Tree LearningIssues in Decision-Tree Learning.

Solving Problems by searching Well defined problems A probem is well defined if it is easy to automatically asses the validity (utility) of any proposed.

1 Universidad de Buenos Aires Maestría en Data Mining y Knowledge Discovery Aprendizaje Automático 5-Inducción de árboles de decisión (2/2) Eduardo Poggi.

Dynamic Programming Discrete time frame Multi-stage decision problem Solves backwards.

Problem Reduction So far we have considered search strategies for OR graph. In OR graph, several arcs indicate a variety of ways in which the original.

Primbs1 Receding Horizon Control for Constrained Portfolio Optimization James A. Primbs Management Science and Engineering Stanford University (with Chang.

Finite State Machines (FSM) OR Finite State Automation (FSA) - are models of the behaviors of a system or a complex object, with a limited number of defined.

Leaving the Script Justin Permar Soar Workshop June 5, 2013 ADAM GT 1

SOAR A cognitive architecture By: Majid Ali Khan.

Beyond Chunking: Learning in Soar March 22, 2003 John E. Laird Shelley Nason, Andrew Nuxoll and a cast of many others University of Michigan.

1 Learning through Interactive Behavior Specifications Tolga Konik CSLI, Stanford University Douglas Pearson Three Penny Software John Laird University.

Problem Solving and Decision Making Intelligence.

Competence-Preserving Retention of Learned Knowledge in Soar’s Working and Procedural Memories Nate Derbinsky, John E. Laird University of Michigan.

Collaborative Decision Making (CDM) Saulo Da Silva

Knowledge Representation and Reasoning

Cognitive Language Processing for Rosie

Learning Fast and Slow John E. Laird

Data Structures and Algorithms

Rational Decision Making 8-step Process

Knowledge Representation and Reasoning

Chapter 3 Component Reliability Analysis of Structures.

Issues in Decision-Tree Learning Avoiding overfitting through pruning

Achieving parsimony between NGS and the Michigan approach

دانشگاه آزاد اسلامی واحد زاهدان

John E. Laird 32nd Soar Workshop

Eight-QueenS Problem: Fuzzy Constraint Satisfaction Problem Using Soar

Bryan Stearns University of Michigan Soar Workshop - May 2018

Learning the Task Definition of Games through ITL

SOAR as a Cognitive Architecture for Modeling Driver Workload

Symbolic cognitive architectures

2 Understanding Variables and Solving Equations.

Introduction Solving inequalities is similar to solving equations. To find the solution to an inequality, use methods similar to those used in solving.

EA C461 – Artificial Intelligence Problem Solving Agents

You replace it with its simplest name

Structure of the final report

Using the Order of Operations

CSE 1342 Programming Concepts

Using the Order of Operations

Presentation transcript:

1 Chunking with Confidence John Laird University of Michigan June 17, th Soar Workshop

2 2 2 Basics of Chunking Compile processing in subgoal into a single rule Rule will fire in future and make subgoal unnecessary.

3 3 3 Basics of Chunking substate superstate substate superstate substate superstate substate result

4 4 4 Original Philosophy Behind Chunking Chunks cache the processing in a subgoal Replace problem solving with recognition Originally assumed goal test was complete Any path from initial state to goal was valid Could ignore control knowledge - just impacts efficiency

5 5 5 Changes in Soar Control knowledge is used to constrain behavior Goal test may be simplified and “assume” control knowledge Results are returned immediately when produced No guaranteed that they fulfill goal Wandering around looking for food – random decisions in subgoal Wander left straightrightstraight

6 6 6 Proposal: Symbolic Preferences Include in chunks the rules that create preferences: they incorporate the “goal concept”. Not just operator proposals and applications Impact: Some chunks will be more specific. Will not have any impact on selection space chunks. Data chunking will be more difficult.

7 7 7 Proposal: Numeric Indifferent Preferences Create chunks only if there is high confidence in result. Result confidence = confidence in each selected operator for every indifferent decisions on path to result Include conditions of rules computing expected value Confidence determined by reinforcement learning Decayed trailing standard deviation of best choice Impact: Chunking will not be immediate when there is indifference Time to chunk will vary Won’t chunk over decisions with no clear best choice Can overcome by deliberately converting to symbolic preferences Chunking “freezes” high confidence solutions

8 8 8 Nuggets and Coal Nugget: Resolves final issues in chunking Could lead to the promised land = ubiquitous chunking Coal: Not implemented Depends on implementation of reinforcement learning Must have rewards = goal tests in subgoals