0% found this document useful (0 votes)
144 views

Lecture 2: Problem Solving Using State Space Representation

The document discusses problem solving using state space representation. It covers key concepts of state space, operators, start and goal states. Search algorithms like depth-first search and breadth-first search are used to traverse the search space represented as a search tree. Different types of problems are formulated using states, operators, and path costs. Examples like Romania, 8-queens, sliding tiles are presented to illustrate state space formulation.

Uploaded by

Vardhan Ammineni
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
144 views

Lecture 2: Problem Solving Using State Space Representation

The document discusses problem solving using state space representation. It covers key concepts of state space, operators, start and goal states. Search algorithms like depth-first search and breadth-first search are used to traverse the search space represented as a search tree. Different types of problems are formulated using states, operators, and path costs. Examples like Romania, 8-queens, sliding tiles are presented to illustrate state space formulation.

Uploaded by

Vardhan Ammineni
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
You are on page 1/ 37

Lecture 2: Problem

Solving using
State Space
Representation
CS 271: Fall, 1006
Overview
 Intelligent agents: problem solving as search

 Search consists of
 state space
 operators
 start state
 goal states
 The search graph

 A Search Tree is an effective way to represent the search


process

 There are a variety of search algorithms, including


 Depth-First Search
 Breadth-First Search
 Others which use heuristic knowledge (in future lectures)
Problem-Solving Agents
 Intelligent agents can solve problems by searching a state-space

 State-space Model
 the agent’s model of the world
 usually a set of discrete states
 e.g., in driving, the states in the model could be towns/cities

 Goal State(s)
 a goal is defined as a desirable state for an agent
 there may be many states which satisfy the goal test
 e.g., drive to a town with a ski-resort
 or just one state which satisfies the goal
 e.g., drive to Mammoth

 Operators (actions, successor function)


 operators are legal actions which the agent can take to move from one
state to another
Example: Romania
 On holiday in Romania; currently in Arad.
 Flight leaves tomorrow from Bucharest
 Formulate goal:
 be in Bucharest
 Formulate problem:
 states: various cities
 actions: drive between cities
 Find solution:
 sequence of cities, e.g., Arad, Sibiu, Fagaras, Bucharest
Example: Romania
Problem types

 Static / Dynamic
Previous problem was static: no attention to changes in environment
 Observable / Partially Observable / Unobservable
Previous problem was observable: it knew its initial state.
 Deterministic / Stochastic
Previous problem was deterministic: no new percepts
were necessary, we can predict the future perfectly
 Discrete / continuous
Previous problem was discrete: we can enumerate all possibilities
State-Space
Problem Formulation
A problem is defined by four items:

initial state e.g., "at Arad“

actions or successor function S(x) = set of action–state pairs


 e.g., S(Arad) = {<Arad  Zerind, Zerind>, … }

goal test, (or goal state)


e.g., x = "at Bucharest”, Checkmate(x)

path cost (additive)


 e.g., sum of distances, number of actions executed, etc.
 c(x,a,y) is the step cost, assumed to be ≥ 0

A solution is a sequence of actions leading from the initial state to a


goal state
Defining Search Problems

 A statement of a Search problem has 4 components


 1. A set of states
 2. A set of “operators” which allow one to get from one state to another
 3. A start state S
 4. A set of possible goal states, or ways to test for goal states
 4a. Cost path

 Search solution consists of


 a sequence of operators which transform S into a goal state G

 Representing real problems in a search framework


 may be many ways to represent states and operators
 key idea: represent only the relevant aspects of the problem (abstraction)
Abstraction
Process of removing irrelevant detail to create an abstract
representation: ``high-level”, ignores irrelevant details

 Definition of Abstraction:
 Navigation Example: how do we define states and operators?
 First step is to abstract “the big picture”
 i.e., solve a map problem
 nodes = cities, links = freeways/roads (a high-level description)
 this description is an abstraction of the real problem
 Can later worry about details like freeway onramps, refueling, etc

 Abstraction is critical for automated problem solving


 must create an approximate, simplified, model of the world for the computer
to deal with: real-world is too detailed to model exactly
 good abstractions retain all important details
Robot block world
 Given a set of block in a certain configuration,
 Move the blocks into a goal configuration.
 Example :
 (c,b,a)  (b,c,a)

A A Move (x,y)
B C
C B
Operator Description
The state-space graph

 Graphs:
 nodes, arcs, directed arcs, paths

 Search graphs:
 States are nodes

 operators are directed arcs

 solution is a path from start to goal

 Problem formulation:
 Give an abstract description of states, operators, initial state and
goal state.
 Problem solving:
 Generate a part of the search space that contains a solution
The Traveling Salesperson
Problem

 Find the shortest tour that visits all cities


without visiting any city twice and return to
starting point.
C
 State: sequence of cities visited
 S0 = A B
A D
F

E
The Traveling Salesperson
Problem

 Find the shortest tour that visits all cities


without visiting any city twice and return to
starting point.
C
 State: sequence of cities visited
 S0 = A B
A D
F

 SG = a complete tour E

{a, c, d } {( a, c, d , x) | X  a, c, d }
Example: 8-queen problem
Example: 8-Queens

 states? -any arrangement of n<=8 queens


-or arrangements of n<=8 queens in leftmost n
columns, 1 per column, such that no queen
attacks any other.
 initial state? no queens on the board
 actions? -add queen to any empty square
-or add queen to leftmost empty square such
that it is not attacked by other queens.
 goal test? 8 queens on the board, none attacked.
 path cost? 1 per move
The sliding tile problem
The Sliding Tile Problem

move( x, loc y, loc z ) Up


Down
Left
Right
The “8-Puzzle” Problem
Start State

1 2 3
4 6
7 5 8

1 2 3
4 5 6
7 8

1 2 3
4 5 6
7 8
Goal State
Abstraction
Process of removing irrelevant detail to create an abstract
representation: ``high-level”, ignores irrelevant details

 Definition of Abstraction:
 Navigation Example: how do we define states and operators?
 First step is to abstract “the big picture”
 i.e., solve a map problem
 nodes = cities, links = freeways/roads (a high-level description)
 this description is an abstraction of the real problem
 Can later worry about details like freeway onramps, refueling, etc

 Abstraction is critical for automated problem solving


 must create an approximate, simplified, model of the world for the computer
to deal with: real-world is too detailed to model exactly
 good abstractions retain all important details
Formulating Problems

 Problem types
 Satisficing: 8-queen
 Optimizing: Traveling salesperson
 Object sought
 board configuration,
 sequence of moves
 A strategy (contingency plan)
 Satisficing leads to optimizing since “small is quick”
 For traveling salesperson
 satisficing easy, optimizing hard
 semi-optimizing:
 Find a good solution
 In Russel and Norvig:
 signle-state, multiple states, contingency plans, exploration problems
Searching the State Space

 States, operators, control strategies


 The search space graph is implicit

 The control strategy generates a small search tree.

 Systematic search
 Do not leave any stone unturned
 Efficiency
 Do not turn any stone more than once
Tree search example
Tree search example
Tree search example
Implementation: states vs. nodes

 A state is a (representation of) a physical configuration


 A node is a data structure constituting part of a search tree
contains info such as: state, parent node, action, path cost g(x),
depth

 The Expand function creates new nodes, filling in the various


fields and using the SuccessorFn of the problem to create the
corresponding states.
Tree Representation of
Searching a State Space
 1. State Space
 nodes are states we can visit
 links are legal transitions between states

 2. Search Tree:
 S is the root node
 The search algorithm searches by expanding leaf nodes
 Internal nodes are states the algorithm has already explored
 Leaves are potential goal nodes: the algorithm stops expanding once it finds
the first goal node G

 Key Concept
 Search trees are a data structure to represent how the search
algorithm explores the state space, i.e.., they dynamically evolve as the
search proceeds
Explicit Graph
Breadth-first of explicit graph
State space of the 8 puzzle
problem
Why Search can be difficult

 At the start of the search, the search algorithm does not know
 the size of the tree
 the shape of the tree
 the depth of the goal states

 How big can a search tree be?


 say there is a constant branching factor b
 and one goal exists at depth d
 search tree which includes a goal can have
bd different branches in the tree (worst case)

 Examples:
 b = 2, d = 10: bd = 210= 1024
 b = 10, d = 10: bd = 1010= 10,000,000,000
A Water Jug Problem
Puzzle-Solving as Search

 You have a 4-gallon and a 3-gallon water jug


 You have a faucet with an unlimited amount of water
 You need to get exactly 2 gallons in 4-gallon jug

 State representation: (x, y)


 x: Contents of four gallon
 y: Contents of three gallon

 Start state: (0, 0)


 Goal state (2, n)
 Operators
 Fill 3-gallon from faucet, fill 4-gallon from faucet
 Fill 3-gallon from 4-gallon , fill 4-gallon from 3-gallon
 Empty 3-gallon into 4-gallon, empty 4-gallon into 3-gallon
 Dump 3-gallon down drain, dump 4-gallon down drain
Production Rules for the Water
Jug Problem
Fill the 4-gallon jug
1 (x,y)  (4,y)
if x < 4 Fill the 3-gallon jug
2 (x,y)  (x,3)
if y < 3 Pour some water out of the 4-gallon jug
3 (x,y)  (x – d,y) Pour some water out of the 3-gallon jug
if x > 0 Empty the 4-gallon jug on the ground
4 (x,y)  (x,y – d)
if x > 0 Empty the 3-gallon jug on the ground
5 (x,y)  (0,y) Pour water from the 3-gallon jug into
if x > 0 the 4-gallon jug until the 4-gallon jug
6 (x,y)  (x,0) is full
if y > 0
7 (x,y)  (4,y – (4 – x))
if x + y ≥ 4 and y > 0
The Water Jug Problem
(cont’d)
8 (x,y)  (x – (3 – y),3) Pour water from the 4-gallon jug
if x + y ≥ 3 and x > 0 into the 3-gallon jug until the
3-gallon jug is full
9 (x,y)  (x + y, 0) Pour all the water from the 3-
gallon jug into the 4-gallon jug
if x + y ≤ 4 and y > 0
Pour all the water from the 4-
10 (x,y)  (0, x + y) gallon jug into the 3-gallon jug
if x + y ≤ 3 and x > 0
Pour the 2 gallons from the 3-
11 (0,2)  (2,0) gallon jug into the 4-gallon jug

12 (x,2)  (0,2) Empty the 4-gallon jug on the


ground
One Solution to the Water Jug
Problem

Gallons in the Gallons in Rule


4-Gallon Jug the 3-Gallon Applied
Jug
0 0 2
0 3 9
3 0 2
3 3 7
4 2 5 or 12
0 2 9 or 11
2 0
Summary
 Intelligent agents can often be viewed as searching for problem
solutions in a discrete state-space

 Search consists of
 state space
 operators
 start state
 goal states

 A Search Tree is an efficient way to represent a search

 There are a variety of general search techniques, including


 Depth-First Search
 Breadth-First Search
 we will look at several others in the next few lectures

 Assigned Reading: Nillson chapter 7 chapter 8


 R&N Chapter 3

You might also like