Artificial Intelligence

Problem Formulation &

Solving Problems by Searching
ENG P KADEBU
 Task environments
2

To design a rational agent we need to specify a task

environment
 a problem specification for which the agent is a solution
PEAS: to specify a task environment
 Performance measure
 Environment
 Actuators
 Sensors
 PEAS: Specifying an automated taxi driver
3

Performance measure:
 safe, fast, legal, comfortable, maximize profits
Environment:
 roads, other traffic, pedestrians, customers
Actuators:
 steering, accelerator, brake, signal, horn
Sensors:
 cameras, sonar, speedometer, GPS
 PEAS: Medical diagnosis system
4

Performance measure:
 Healthy patient, minimize costs, lawsuits
Environment:
 Patient, hospital, staff
Actuators:
 Screen display (form including: questions, tests, diagnoses,
treatments, referrals)
Sensors:
 Keyboard (entry of symptoms, findings, patient's answers)
 Environment Restrictions for Now
5

Static
Fully Observable
Deterministic
Discrete
 The rational agent designer’s goal
6

Goal of AI practitioner who designs rational agents:

given a PEAS task environment,
1.Construct agent function f that maximizes (the expected value
of) the performance measure,
2.Design an agent program that implements f on a particular
architecture
 Problem Solving Agents
7

Problem solving agent

 A kind of “goal based” agent
 Finds sequences of actions that lead to desirable states (goal).

The algorithms are uninformed

 No extra information about the problem other than the
definition
 No extra information
 No heuristics (rules)
 Goal Based Agent
8

What the world Sensors Percepts

State is like now

Environment
How the world evolves

What my actions do
What it will be like
if I do action A

What action I
Goals Actuators Actions
should do now
 Goal Based Agents
9

 Assumes the problem environment is:

 Static
 The plan remains the same
 Observable
 Agent knows the initial state
 Discrete
 Agent can enumerate the choices
 Deterministic
 Agent can plan a sequence of actions such that each will lead to an
intermediate state

 The agent carries out its plans with its eyes closed
 Certain of what’s going on
 Open loop system
 Well Defined Problems and Solutions
10

A problem
 Initial state
 Actions and Successor Function
 Goal test
 Path cost
 Problem formulation
11

Choosing a relevant set of states to consider, and a

feasible set of operators for moving from one state to
another.
Search is the process of considering various possible
sequences of operators applied to the initial state,
and finding out a sequence which culminates in a
goal state.
 Four general steps in problem solving
12

 Goal formulation
 Deciding on goal states based on current situation and agent’s
performance measure
 Problem formulation
 how can we get to the goal, without getting bogged down in the detail
of the world.
 What actions and states to consider given the goal
 Search
 Determine the possible sequence of actions that lead to the states of
known values and then choose the best sequence.
 Search algorithms –input is a problem, output is a solution (action
sequence)
 Execute–Given the solution, perform the actions.
 State Space Search Notations
13

 An initial state is the description of the starting

configuration of the agent
 An action or an operator takes the agent from one state to
another state which is called a successor state.
 A state can have a number of successor states.
 A plan is a sequence of actions.
 The cost of a plan is referred to as the path cost. The path
cost is a positive number, and a common path cost may be
the sum of the costs of the steps in the path.
[S,s,O,G]
 Search Problem
14

 We are now ready to formally describe a search

problem.
A search problem consists of the following:
• S: the full set of states
• s : the initial state
• A:S→S is a set of operators
• G is the set of final states.
Note that G ⊆S
State Space Search
15
 Representation of search problems
16

 A search problem is represented using a directed

graph.
The states are represented as nodes.
The allowed actions are represented as arcs.
 Searching process
17

 The generic searching process can be very simply

described in terms of the following steps:
Do until a solution is found or the state space is
exhausted.
1. Check the current state
2. Execute allowable actions to find the successor states.
3. Pick one of the new states.
4. Check if the new state is a solution state If it is not, the new
state becomes the current state and the process is repeated
Illustration
18
Goal State found
19
8-Puzzle problem
20
 State Space
21

The state space representation for this problem is

summarized below:
 States: A state is a description of each of the

eight tiles in each location that it can occupy.

 Operators/Action: The blank moves left, right, up
or down
 Goal Test: The current state matches a certain
state (e.g. one of the ones shown on previous
slide)
 Path Cost: Each move of the blank costs 1
A portion of the 8-puzzle
22
 Example: Water Jug Problem
23

Given a 4 gallon bucket and a 3 gallon bucket, how

can we measure exactly 2 gallons into one bucket?
 There are no markings on the bucket
 You must fill each bucket completely
 Example: Water Jug Problem
24

Initial state:
 The buckets are empty
 Represented by the tuple ( 0 0 )

Goal state:
 One of the buckets has two gallons of water in it
 Represented by either ( x 2 ) or ( 2 y )

Path cost:
 1 per unit step
 Example: Water Pouring
25

Actions and Successor Function

 Fill a bucket
 (x y) -> (3 y)
 (x y) -> (x 4)
 Empty a bucket
 (x y) -> (0 y)
 (x y) -> (x 0)
 Pour contents of one bucket into another
 (x y) -> (0 x+y) or (x+y-4, 4)
 (x y) -> (x+y 0) or (3, x+y-3)
Example: Water Pouring
26

(0,0)

(4,0) (0,3)

(1,3) (4,3) (3,0)

(1,0) (0,1)
(3,3) (4,2)

(4,1)
(2,3)

(2,0) (0,2)
 Example: Romania
27

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

Semester 2 , 2018
Example: Map Planning
28
Well-defined problems and solutions
29
 Searching For Solutions
30

Initial State
 e.g. “At Arad”
Successor Function
 A set of action state pairs
 S(Arad) = {(Arad->Zerind, Zerind), …}
Goal Test
 e.g. x = “at Bucharest”
Path Cost
 sum of the distances traveled
Problem Formulation
31
 Searching For Solutions
32

Having formulated some problems…how do we solve

them?

Search through a state space

Use a search tree that is generated with an initial

state and successor functions that define the state
space
 Searching For Solutions
33

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

 A node is a data structure constituting part of a search tree

 Includes parent, children, depth, path cost

 States do not have children, depth, or path cost

 The EXPAND function creates new nodes, filling in the

various fields and using the SUCCESSOR function of the
problem to create the corresponding states
Searching For Solutions
34
 Implementation: Components of a node
35

State: the state in the state space to which the node

corresponds
Parent-node: the node in the search tree that generated
this node
Action: the action that was applied to the parent to
generate the node
Path-cost: the cost, traditionally denoted by g(n), of the
path from the initial state to the node, as indicated by the
parent pointers
Depth: the number of steps along the path from the initial
state
Semester 2 , 2018
Searching For Solutions
36
Thank You!!
37