cs460 spring 2011
DESCRIPTION
CS460 Spring 2011. Lecture 2. Review. AI definition: rational agent [ f : P* A ] Agent Model: PEAS Environment Types Agent Types Vacuum cleaner world. Environment types. Chess with Chess without Taxi driving a clocka clock Fully observableYesYesNo - PowerPoint PPT PresentationTRANSCRIPT
![Page 1: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/1.jpg)
CS460 Spring 2011
Lecture 2
22/04/23 1
![Page 2: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/2.jpg)
Review AI definition: rational agent [f: P* A] Agent Model: PEAS Environment Types Agent Types Vacuum cleaner world
22/04/23 2
![Page 3: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/3.jpg)
Environment typesChess with Chess without Taxi
driving a clock a clock
Fully observable Yes Yes No Deterministic Strategic Strategic No Episodic No No No Static Semi Yes No Discrete Yes Yes NoSingle agent No No No
The environment type largely determines the agent design The real world is (of course) partially observable, stochastic,
sequential, dynamic, continuous, multi-agent
![Page 4: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/4.jpg)
Agent types Four basic types in order of increasing
generality: Simple reflex agents Model-based reflex agents Goal-based agents Utility-based agents
![Page 5: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/5.jpg)
Simple reflex agents
![Page 6: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/6.jpg)
Model-based reflex agents
![Page 7: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/7.jpg)
Model-based reflex agents
![Page 8: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/8.jpg)
Utility-based agents
![Page 9: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/9.jpg)
Utility-based agents
![Page 10: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/10.jpg)
Vacuum-cleaner world
Percepts: location and contents, e.g., [A,Dirty]
Actions: Left, Right, Suck, NoOp
![Page 11: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/11.jpg)
22/04/23 11
Solving problems by searching
Chapter 3
![Page 12: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/12.jpg)
22/04/23 12
Outline Problem-solving agents Problem types Problem formulation Example problems Basic search algorithms
![Page 13: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/13.jpg)
22/04/23 13
Problem-solving agents
![Page 14: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/14.jpg)
22/04/23 14
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
![Page 15: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/15.jpg)
22/04/23 15
Example: Romania
![Page 16: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/16.jpg)
22/04/23 16
Problem types Deterministic, fully observable single-state
problem Agent knows exactly which state it will be in; solution is a
sequence Non-observable sensorless problem (conformant
problem) Agent may have no idea where it is; solution is a sequence
Nondeterministic and/or partially observable contingency problem
percepts provide new information about current state often interleave} search, execution
Unknown state space exploration problem
![Page 17: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/17.jpg)
22/04/23 17
Example: vacuum world Single-state, start in
#5. Solution?
![Page 18: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/18.jpg)
22/04/23 18
Example: vacuum world Single-state, start in #5.
Solution? [Right, Suck]
Sensorless, start in {1,2,3,4,5,6,7,8} e.g., Right goes to {2,4,6,8} Solution?
![Page 19: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/19.jpg)
22/04/23 19
Example: vacuum world Sensorless, start in
{1,2,3,4,5,6,7,8} e.g., Right goes to {2,4,6,8} Solution? [Right,Suck,Left,Suck]
Contingency Nondeterministic: Suck may
dirty a clean carpet Partially observable: location, dirt at current location. Percept: [L, Clean], i.e., start in #5 or #7
Solution?
![Page 20: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/20.jpg)
22/04/23 20
Example: vacuum world Sensorless, start in
{1,2,3,4,5,6,7,8} e.g., Right goes to {2,4,6,8} Solution? [Right,Suck,Left,Suck]
Contingency Nondeterministic: Suck may
dirty a clean carpet Partially observable: location, dirt at current location. Percept: [L, Clean], i.e., start in #5 or #7
Solution? [Right, if dirt then Suck]
![Page 21: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/21.jpg)
22/04/23 21
Single-state problem formulation
A problem is defined by four items:
1. initial state e.g., "at Arad"2. actions or successor function S(x) = set of action–state pairs
e.g., S(Arad) = {<Arad Zerind, Zerind>, … }3. goal test, can be
explicit, e.g., x = "at Bucharest" implicit, e.g., Checkmate(x)
n path cost (additive)1. e.g., sum of distances, number of actions executed, etc.2. c(x,a,y) is the step cost, assumed to be ≥ 0
1. A solution is a sequence of actions leading from the initial state to a goal state
2.1.
![Page 22: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/22.jpg)
22/04/23 22
Selecting a state space Real world is absurdly complex
state space must be abstracted for problem solving (Abstract) state = set of real states (Abstract) action = complex combination of real
actions e.g., "Arad Zerind" represents a complex set of possible
routes, detours, rest stops, etc. For guaranteed realizability, any real state "in Arad“
must get to some real state "in Zerind" (Abstract) solution =
set of real paths that are solutions in the real world Each abstract action should be "easier" than the
original problem
![Page 23: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/23.jpg)
22/04/23 23
Vacuum world state space graph
states? actions? goal test? path cost?
![Page 24: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/24.jpg)
22/04/23 24
Vacuum world state space graph
states? integer dirt and robot location actions? Left, Right, Suck goal test? no dirt at all locations path cost? 1 per action
![Page 25: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/25.jpg)
22/04/23 25
Example: The 8-puzzle
states? actions? goal test? path cost?
![Page 26: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/26.jpg)
22/04/23 26
Example: The 8-puzzle
states? locations of tiles actions? move blank left, right, up, down goal test? = goal state (given) path cost? 1 per move
[Note: optimal solution of n-Puzzle family is NP-hard]
![Page 27: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/27.jpg)
22/04/23 27
Example: robotic assembly
states?: real-valued coordinates of robot joint angles parts of the object to be assembled
actions?: continuous motions of robot joints goal test?: complete assembly path cost?: time to execute
![Page 28: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/28.jpg)
22/04/23 28
Tree search algorithms Basic idea:
offline, simulated exploration of state space by generating successors of already-explored states (a.k.a.~expanding states)
(from 2nd ed)
(from
![Page 29: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/29.jpg)
22/04/23 29
Tree search example
![Page 30: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/30.jpg)
22/04/23 30
Tree search example
![Page 31: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/31.jpg)
22/04/23 31
Tree search example
![Page 32: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/32.jpg)
22/04/23 32
Implementation: general tree search
![Page 33: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/33.jpg)
22/04/23 33
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 includes 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.
![Page 34: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/34.jpg)
22/04/23 34
Search strategies A search strategy is defined by picking the order of
node expansion Strategies are evaluated along the following
dimensions: completeness: does it always find a solution if one exists? time complexity: number of nodes generated space complexity: maximum number of nodes in memory optimality: does it always find a least-cost solution?
Time and space complexity are measured in terms of
b: maximum branching factor of the search tree d: depth of the least-cost solution m: maximum depth of the state space (may be ∞)
![Page 35: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/35.jpg)
22/04/23 35
Uninformed search strategies
Uninformed search strategies use only the information available in the problem definition
Breadth-first search Uniform-cost search Depth-first search Depth-limited search Iterative deepening search
![Page 36: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/36.jpg)
22/04/23 36
Breadth-first search Expand shallowest unexpanded node Implementation:
fringe is a FIFO queue, i.e., new successors go at end
![Page 37: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/37.jpg)
22/04/23 37
Breadth-first search Expand shallowest unexpanded node Implementation:
fringe is a FIFO queue, i.e., new successors go at end
![Page 38: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/38.jpg)
22/04/23 38
Breadth-first search Expand shallowest unexpanded node Implementation:
fringe is a FIFO queue, i.e., new successors go at end
![Page 39: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/39.jpg)
22/04/23 39
Breadth-first search Expand shallowest unexpanded node Implementation:
fringe is a FIFO queue, i.e., new successors go at end
![Page 40: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/40.jpg)
22/04/23 40
Properties of breadth-first search
Complete? Yes (if b is finite) Time? 1+b+b2+b3+… +bd + b(bd-1) =
O(bd+1) Space? O(bd+1) (keeps every node in
memory) Optimal? Yes (if cost = 1 per step)
Space is the bigger problem (more than time)
![Page 41: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/41.jpg)
22/04/23 41
Uniform-cost search Expand least-cost unexpanded node Implementation:
fringe = queue ordered by path cost (priority queue) Equivalent to breadth-first if step costs all equal Complete? Yes, if step cost ≥ ε Time? # of nodes with g ≤ cost of optimal solution,
O(bceiling(C*/ ε)) where C* is the cost of the optimal solution
Space? # of nodes with g ≤ cost of optimal solution, O(bceiling(C*/ ε))
Optimal? Yes – nodes expanded in increasing order of g(n)
![Page 42: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/42.jpg)
22/04/23 42
Depth-first search Expand deepest unexpanded node Implementation:
fringe = LIFO queue, i.e., put successors at front
![Page 43: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/43.jpg)
22/04/23 43
Depth-first search Expand deepest unexpanded node Implementation:
fringe = LIFO queue, i.e., put successors at front
![Page 44: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/44.jpg)
22/04/23 44
Depth-first search Expand deepest unexpanded node Implementation:
fringe = LIFO queue, i.e., put successors at front
![Page 45: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/45.jpg)
22/04/23 45
Depth-first search Expand deepest unexpanded node Implementation:
fringe = LIFO queue, i.e., put successors at front
![Page 46: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/46.jpg)
22/04/23 46
Depth-first search Expand deepest unexpanded node Implementation:
fringe = LIFO queue, i.e., put successors at front
![Page 47: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/47.jpg)
22/04/23 47
Depth-first search Expand deepest unexpanded node Implementation:
fringe = LIFO queue, i.e., put successors at front
![Page 48: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/48.jpg)
22/04/23 48
Depth-first search Expand deepest unexpanded node Implementation:
fringe = LIFO queue, i.e., put successors at front
![Page 49: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/49.jpg)
22/04/23 49
Depth-first search Expand deepest unexpanded node Implementation:
fringe = LIFO queue, i.e., put successors at front
![Page 50: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/50.jpg)
22/04/23 50
Depth-first search Expand deepest unexpanded node Implementation:
fringe = LIFO queue, i.e., put successors at front
![Page 51: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/51.jpg)
22/04/23 51
Depth-first search Expand deepest unexpanded node Implementation:
fringe = LIFO queue, i.e., put successors at front
![Page 52: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/52.jpg)
22/04/23 52
Depth-first search Expand deepest unexpanded node Implementation:
fringe = LIFO queue, i.e., put successors at front
![Page 53: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/53.jpg)
22/04/23 53
Depth-first search Expand deepest unexpanded node Implementation:
fringe = LIFO queue, i.e., put successors at front
![Page 54: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/54.jpg)
22/04/23 54
Properties of depth-first search
Complete? No: fails in infinite-depth spaces, spaces with loops Modify to avoid repeated states along path
complete in finite spaces Time? O(bm): terrible if m is much larger than
d but if solutions are dense, may be much faster
than breadth-first Space? O(bm), i.e., linear space! Optimal? No
![Page 55: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/55.jpg)
22/04/23 55
Depth-limited search= depth-first search with depth limit l,i.e., nodes at depth l have no successors
Recursive implementation:
![Page 56: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/56.jpg)
22/04/23 56
Iterative deepening search
![Page 57: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/57.jpg)
22/04/23 57
Iterative deepening search l =0
![Page 58: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/58.jpg)
22/04/23 58
Iterative deepening search l =1
![Page 59: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/59.jpg)
22/04/23 59
Iterative deepening search l =2
![Page 60: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/60.jpg)
22/04/23 60
Iterative deepening search l =3
![Page 61: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/61.jpg)
22/04/23 61
Iterative deepening search Number of nodes generated in a depth-limited search to
depth d with branching factor b: NDLS = b0 + b1 + b2 + … + bd-2 + bd-1 + bd
Number of nodes generated in an iterative deepening search to depth d with branching factor b:
NIDS = (d+1)b0 + d b^1 + (d-1)b^2 + … + 3bd-2 +2bd-1 + 1bd
For b = 10, d = 5, NDLS = 1 + 10 + 100 + 1,000 + 10,000 + 100,000 = 111,111 NIDS = 6 + 50 + 400 + 3,000 + 20,000 + 100,000 = 123,456
Overhead = (123,456 - 111,111)/111,111 = 11%
![Page 62: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/62.jpg)
22/04/23 62
Properties of iterative deepening search
Complete? Yes Time? (d+1)b0 + d b1 + (d-1)b2 + … +
bd = O(bd) Space? O(bd) Optimal? Yes, if step cost = 1
![Page 63: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/63.jpg)
22/04/23 63
Summary of algorithms
![Page 64: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/64.jpg)
22/04/23 64
Repeated states Failure to detect repeated states can
turn a linear problem into an exponential one!
![Page 65: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/65.jpg)
22/04/23 65
Graph search
![Page 66: CS460 Spring 2011](https://reader035.vdocuments.net/reader035/viewer/2022062520/56816345550346895dd3d4ff/html5/thumbnails/66.jpg)
22/04/23 66
Summary Problem formulation usually requires abstracting
away real-world details to define a state space that can feasibly be explored
Variety of uninformed search strategies
Iterative deepening search uses only linear space and not much more time than other uninformed algorithms