Classical Planning - Artificial Intelligence - Lecture Slides, Slides of Artificial Intelligence

Download Classical Planning - Artificial Intelligence - Lecture Slides and more Slides Artificial Intelligence in PDF only on Docsity!

Lecture 19 of 41

Introduction to Classical Planning

Lecture Outline

• Today’s Reading
• Friday’s Reading: Sections 11.5 – 11.9, Russell and Norvig
• Previously: Logical Representations and Theorem Proving
  • Propositional, predicate, and first-order logical languages
  • Proof procedures: forward and backward chaining, resolution refutation
• Today: Introduction to Classical Planning
  • Search vs. planning
  • STRIPS axioms
• Wednesday: More Classical Planning
  • Partial-order planning (NOAH, etc.)
  • Limitations
• First Hour Exam: Wednesday, 13 Oct 2004
  • Remote students: have exam agreement faxed to DCE
  • Exam will be faxed to proctors Wednesday or Friday

Planning in Situation Calculus

STRIPS Operators

Midterm Review – IAs, Search:

Unclear Points?

• Artificial Intelligence (AI)
  • Operational definition: study / development of systems capable of “thought processes” (reasoning, learning, problem solving)
  • Constructive definition: expressed in artifacts (design and implementation)
• Intelligent Agent Framework
  • Reactivity vs. state
  • From goals to preferences (utilities)
• Methodologies and Applications
  • Search: game-playing systems, problem solvers
  • Planning, design, scheduling systems
  • Control and optimization systems
  • Machine learning: hypothesis space search (for pattern recognition, data mining)
• Search
  • Problem formulation: state space (initial / operator / goal test / cost), graph
  • State space search approaches
    • Blind (uninformed) – DFS, BFS, B&B
    • Heuristic (informed) – Greedy, Beam, A/A; Hill-Climbing, SA*

Midterm Review – Game Trees:

Unclear Points?

• Games as Search Problems
  • Frameworks
  • Concepts: utility, reinforcements, game trees
  • Static evaluation under resource limitations
• Family of Algorithms for Game Trees: Minimax
  • Static evaluation algorithm
    • To arbitrary ply
    • To fixed ply
    • Sophistications: iterative deepening, alpha-beta pruning
  • Credit propagation
    • Intuitive concept
    • Basis for simple (delta-rule) learning algorithms
• State of The Field
• Uncertainty in Games: Expectiminimax and Other Algorithms

Describing Actions [1]:

Frame, Qualification, and Ramification Problems

Adapted from slides by S. Russell, UC Berkeley

Adapted from slides by S. Russell, UC Berkeley

Describing Actions [2]:

Successor State Axioms

Making Plans:

A Better Way

Adapted from slides by S. Russell, UC Berkeley

First-Order Logic:

Summary

Adapted from slides by S. Russell, UC Berkeley

POP Algorithm [1]:

Sketch

Adapted from slides by S. Russell, UC Berkeley

Adapted from slides by S. Russell, UC Berkeley

POP Algorithm [2]:

Subroutines and Properties

Summary Points

• Previously: Logical Representations and Theorem Proving
  • Propositional, predicate, and first-order logical languages
  • Proof procedures: forward and backward chaining, resolution refutation
• Today: Introduction to Classical Planning
  • Search vs. planning
  • STRIPS axioms
    • Operator representation
    • Components: preconditions, postconditions (ADD, DELETE lists)
• Thursday: More Classical Planning
  • Partial-order planning (NOAH, etc.)
  • Limitations

Adapted from slides by S. Russell, UC Berkeley

Terminology

• Classical Planning
  • Planning versus search
  • Problematic approaches to planning
    • Forward chaining
    • Situation calculus
  • Representation
    • Initial state
    • Goal state / test
    • Operators
• Efficient Representations
  • STRIPS axioms
    • Components: preconditions, postconditions (ADD, DELETE lists)
    • Clobbering / threatening
  • Reactive plans and policies
  • Markov decision processes