Heuristics in Search-Space CSE 574 April 11, 2003 Dan Weld.

Slides:



Advertisements
Similar presentations
Artificial Intelligence Presentation
Advertisements

An Introduction to Artificial Intelligence
CLASSICAL PLANNING What is planning ?  Planning is an AI approach to control  It is deliberation about actions  Key ideas  We have a model of the.
PROBLEM SOLVING AND SEARCH
Top 5 Worst Times For A Conference Talk 1.Last Day 2.Last Session of Last Day 3.Last Talk of Last Session of Last Day 4.Last Talk of Last Session of Last.
Finding Search Heuristics Henry Kautz. if State[node] is not in closed OR g[node] < g[LookUp(State[node],closed)] then A* Graph Search for Any Admissible.
Plan Generation & Causal-Link Planning 1 José Luis Ambite.
CSCI 5582 Fall 2006 CSCI 5582 Artificial Intelligence Lecture 4 Jim Martin.
Solving Problem by Searching
Artificial Intelligence Chapter 9 Heuristic Search Biointelligence Lab School of Computer Sci. & Eng. Seoul National University.
1 Heuristic Search Chapter 4. 2 Outline Heuristic function Greedy Best-first search Admissible heuristic and A* Properties of A* Algorithm IDA*
Lecture 3: Uninformed Search
Heuristic State Space Seach Henry Kautz. Assignment.
Problem Solving by Searching
UnInformed Search What to do when you don’t know anything.
MAE 552 – Heuristic Optimization Lecture 27 April 3, 2002
CSE 5731 Lecture 21 State-Space Search vs. Constraint- Based Planning CSE 573 Artificial Intelligence I Henry Kautz Fall 2001.
Handling non-determinism and incompleteness. Problems, Solutions, Success Measures: 3 orthogonal dimensions  Incompleteness in the initial state  Un.
CS 460 Spring 2011 Lecture 3 Heuristic Search / Local Search.
Informed Search CSE 473 University of Washington.
Non-Conservative Cost Bound Increases in IDA* Doug Demyen.
Class of 28 th August. Announcements Lisp assignment deadline extended (will take it until 6 th September (Thursday). In class. Rao away on 11 th and.
Informed Search Idea: be smart about what paths to try.
Classical Planning Chapter 10.
Vilalta&Eick: Informed Search Informed Search and Exploration Search Strategies Heuristic Functions Local Search Algorithms Vilalta&Eick: Informed Search.
Dana Nau: Lecture slides for Automated Planning Licensed under the Creative Commons Attribution-NonCommercial-ShareAlike License:
Informed State Space Search Department of Computer Science & Engineering Indian Institute of Technology Kharagpur.
Informed search algorithms
Midterm Review Prateek Tandon, John Dickerson. Basic Uninformed Search (Summary) b = branching factor d = depth of shallowest goal state m = depth of.
CSE 473: Artificial Intelligence Spring 2012
Informed searching. Informed search Blind search algorithms do not consider any information about the states and the goals Often there is extra knowledge.
Informed Search Include specific knowledge to efficiently conduct the search and find the solution.
Problem Spaces & Search CSE 573. © Daniel S. Weld 2 Logistics Mailing list Reading Ch 4.2, Ch 6 Mini-Project I Partners Game Playing?
1 Branch and Bound Searching Strategies Updated: 12/27/2010.
Heuristic Search Andrea Danyluk September 16, 2013.
For Wednesday Read chapter 6, sections 1-3 Homework: –Chapter 4, exercise 1.
Informed Search I (Beginning of AIMA Chapter 4.1)
Search (continued) CPSC 386 Artificial Intelligence Ellen Walker Hiram College.
Classical Planning Chapter 10 Mausam / Andrey Kolobov (Based on slides of Dan Weld, Marie desJardins)
Informed Search CSE 473 University of Washington.
© Daniel S. Weld 1 Logistics Travel Wed class led by Mausam Week’s reading R&N ch17 Project meetings.
Graphplan CSE 574 April 4, 2003 Dan Weld. Schedule BASICS Intro Graphplan SATplan State-space Refinement SPEEDUP EBL & DDB Heuristic Gen TEMPORAL Partial-O.
CE 473: Artificial Intelligence Autumn 2011 A* Search Luke Zettlemoyer Based on slides from Dan Klein Multiple slides from Stuart Russell or Andrew Moore.
CHAPTER 2 SEARCH HEURISTIC. QUESTION ???? What is Artificial Intelligence? The study of systems that act rationally What does rational mean? Given its.
3.5 Informed (Heuristic) Searches This section show how an informed search strategy can find solution more efficiently than uninformed strategy. Best-first.
Heuristic Search Planners. 2 USC INFORMATION SCIENCES INSTITUTE Planning as heuristic search Use standard search techniques, e.g. A*, best-first, hill-climbing.
Romania. Romania Problem Initial state: Arad Goal state: Bucharest Operators: From any node, you can visit any connected node. Operator cost, the.
CMPT 463. What will be covered A* search Local search Game tree Constraint satisfaction problems (CSP)
Chapter 3.5 Heuristic Search. Learning Objectives Heuristic search strategies –Best-first search –A* algorithm Heuristic functions.
Department of Computer Science
Informed Search Chapter 4 (a)
Discussion on Greedy Search and A*
Discussion on Greedy Search and A*
CSE 4705 Artificial Intelligence
Planning José Luis Ambite.
Artificial Intelligence Informed Search Algorithms
COMP 8620 Advanced Topics in AI
Week 4 Jan 29th 2016.
Artificial Intelligence
Informed search algorithms
CSE 473 University of Washington
Informed Search Idea: be smart about what paths to try.
Search.
SEG 4560 Midterm Review.
Artificial Intelligence
Search.
Informed Search Idea: be smart about what paths to try.
Supplemental slides for CSE 327 Prof. Jeff Heflin
“If I Only had a Brain” Search
Presentation transcript:

Heuristics in Search-Space CSE 574 April 11, 2003 Dan Weld

Schedule 1. BASICS Intro Graphplan SATplan State-space Refinement 2. SPEEDUP EBL & DDB Heuristic Gen 3. TEMPORAL Partial-O Graphplan Forward- chaining Stochastic 4. UNCERTAINTY Long paper!

More Administrivia Mailing List Reviews due by 11am No class Fri 4/18 Experimenting with Planners Context Basis for Projects

Paper: Main Points Avoid duplicate work computing heuristics Pregenerate in forward sweep; search backward Interpretation of graphplan Tradeoff for heuristics: admissible?

Regression Search

Experiments What were they answering? Weak support for main points Do you believe GP analysis? Why is HSPr faster? Heuristic calculation or backwards search? Presentation: Table vs. graph Speedup ratio (how compared to 85%)

Weaknesses Experiments Needed an example More discussion of mutex tradeoffs Algo finds “most” mutexes Memory usage is a problem (but why?) Hill climbing with w=5

Greedy BFS F(n) = g(n) + W h(n) Empirical studies 1979, 1989 on 8 puz, TSP Increasing W => speeds search => less optimal solutions Proof: if W  1 Then |solution|/|optimal|  W Proof: abstract tree, uniform branch, 1 goal Then W=1 gives fastest solution Optimality is a bonus Contradiction?!

Future Work Study graphplan in state-space framework Project idea: Use Hg with IDA* Recast other planners as HSP Analysis of where and why HSPr fails Empirical Comparison A*, IDA*, BFS, and HSPr

Future Work 2 Derivation of better heuristics Keep some delete effects Sum vs. max heuristic Connection to parallel actions Probabilistic estimate of step reuse Non admissible, but more accurate? Can we bound the amount of step sharing?

Bounding Step Sharing Build bipartite “support” graph Compute “max non-mutex outdegree” Good project? A1A1 A2A2 A3A3 A4A4 A1A1 A2A2 A3A3 A4A4 PreEff Mutex

McDermott’s Grid World

Future Work 3 Other points on mutex-computation spectrum HSPr mutexes are limited to pairs of atoms Do not consider actions directly Could ADDs help state enumeration problem

Future Work 4 Progression / Regression Why not generate heuristics backwards? For what domains would this be faster? Advantages? Anytime planning

Future Work 5 Incomplete Info / Temporal What is state space? How to regress?