Download Problem Solving.pps

EA C461 – Artificial Intelligence
Problem Solving Agents
S.P.Vimal
http://discovery.bits-pilani.ac.in/~vimalsp/1910AI/
To discuss…
• Structure of an Agent
• Problem Solving Agents
Vimal
EA C461- Artificial Intelligence
Structure of Agents
Agent = Architecture
+
Computing device Running
Agent Program, with
sensors & actuators
Vimal
Program
Implements Agent Function,
performs mapping of
percepts to actions
EA C461- Artificial Intelligence
Agent Programs
function Table-Driven-Agent (percept) returns an action
static : percepts, a sequence, initially empty
table, a table of actions, indexed by percept sequence, initially fully specified
append percept to the end of percepts
action  LookUp (percepts,table)
return action
• For P percepts, T Life time, Lookup table will
have Σt|P|t entries.
• Large table size
– Need small programs implement rational behavior
Vimal
EA C461- Artificial Intelligence
Agent Programs
• Kinds of Agent Programs
–
–
–
–
Vimal
Simple Reflex Agents
Model-based Reflex Agents
Goal Based Reflex Agents
Utility-based Reflex Agents
EA C461- Artificial Intelligence
Simple Reflex Agents
• Considers only the current percept, ignores rest
of percept history
function Reflex-Vacuum-Agent ([location, status]) returns an action
If status=Dirty then return Suck
else if location=A then return Right
else if location=B return Left
• Condition-action rules encoded
– If car-in-front-is-braking then initiate-braking
function Simple-Reflex-Agent (percept) returns an action
static: rules, a set of condition-action rules
state  Interpret-Input (percept)
rule  Rule-Match (state, rules)
action  Rule-Action [rule]
return action
Vimal
EA C461- Artificial Intelligence
Simple Reflex Agents
rectangles  the current internal state; Ovals  background information
Vimal
EA C461- Artificial Intelligence
Model-Based Reflex Agents
• Keep track of the part of the world which
the agent can’t see now
– Handling partial observability
– Maintain an internal state depends the percept
history
• Updating internal state of the agent needs
some information about
– how the world evolves
– the agent’s own action affects the world
Vimal
EA C461- Artificial Intelligence
Model-Based Reflex Agents
rectangles  the current internal state; Ovals  background information
Vimal
EA C461- Artificial Intelligence
Model-Based Reflex Agents
function Reflex-Agent-With-State (percept) returns an action
static: rules, a set of condition-action rules
state, a description of the current world state
action, the most recent action, initially none
state  Update-Input (state, action, percept)
rule  Rule-Match (state, rules)
action  Rule-Action [rule]
return action
Vimal
EA C461- Artificial Intelligence
Goal-Based Agents
• Having a goal, combined with the current state
information can help select the possible next
action
– Possibly agent may need to consider every alternative
action sequences leading to the goal  search for a
sequence leading to goal
Vimal
EA C461- Artificial Intelligence
Goal-Based Agents
rectangles  the current internal state; Ovals  background information
Vimal
EA C461- Artificial Intelligence
Utility-Based Agents
• Goals provide crude binary distinction between
“happy” and “un happy”
– If one state is preferred over the other, then it has
higher utility for the agent
utility-function (state) = real number (degree of happiness)
• Complete specification of utility-function allows
rational decisions in the following circumstances
– Taking decision when in presence of Conflicting goals
– When there are several goals that the agent can aim for.
Vimal
EA C461- Artificial Intelligence
Utility-Based Agents
rectangles  the current internal state; Ovals  background information
Vimal
EA C461- Artificial Intelligence
Learning Agents
• Build a learning machine and teach it
• Learning agent has the following components
– Learning element
• Suggests modification to the existing rule to the critic
– Performance element
• Collection of knowledge and procedures for selecting the
driving actions
• Choice depends on Learning element
– Critic
• Observes the world and passes information to the learning
element
– Problem generator
• Identifies certain areas of behavior needs improvement and
suggest experiments
Vimal
EA C461- Artificial Intelligence
Learning Agents
rectangles  the current internal state; Ovals  background information
Vimal
EA C461- Artificial Intelligence
Problem Solving Agent
• A kind of Goal based Agent
• Decides what to do by finding the sequences of
actions that lead to desirable states
Formulate Goal, Formulate Problem
Search
Execute
Vimal
EA C461- Artificial Intelligence
Problems
• Four components of problem definition
– Initial state
– Possible Actions
• Uses a Successor Function
– Returns <action, successor> pair
• State Space
• Path
– Goal Test
– Path cost
• Step cost
• Problem formulation is the process of deciding
what actions and states to consider, given a goal
Vimal
EA C461- Artificial Intelligence
Solutions
• A Solution to the problem is the path from the initial state
to the final state
• Quality of solution is measured by path cost function
– Optimal Solution has the lowest path cost among other
solutions
• An Agent with several immediate options of
unknown value can decide what to do by first
examining different possible sequences of actions
that lead to a state of known value, and then
choosing the best sequence  Searching
Process
– Input to Search
– Output from Search
Vimal
: Problem
: Solution in the form of Action Sequence
EA C461- Artificial Intelligence
Problem Solving Agent
A Problem solving
Agent, Assuming
the environment is
• Static
• Observable
• Discrete
• Deterministic
Vimal
EA C461- Artificial Intelligence
Problem Solving Agent : Example
A Simplified Road Map of Part of Romania
Vimal
EA C461- Artificial Intelligence
Problem Solving Agent : Example
• 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
Vimal
EA C461- Artificial Intelligence
Example-1 : Vacuum World
Problem Formulation
• States
– 2 x 22 = 8 states
• Initial State
– Any one of 8 states
• Successor Function
– Legal states that result from three actions
(Left, Right, Suck)
• Goal Test
– All squares are clean
• Path Cost
– Number of steps (each step costs a value
of 1)
Vimal
EA C461- Artificial Intelligence
Example-1 : Vacuum World
State Space for the Vacuum World.
Labels on Arcs denote
Vimal
L: Left, R: Right, S: Suck
EA C461- Artificial Intelligence
Example-2 : The 8-Puzzle
A typical Instance of 8-puzzle
•
•
•
•
•
Vimal
States ?
Initial State ?
Successor Function ?
Goal Test ?
Path Cost ?
EA C461- Artificial Intelligence
Example-2 : The 8-Puzzle
A typical Instance of 8-puzzle
•
•
•
•
•
Vimal
States
: Location of Tiles
Initial State
: One of States
Successor Function: Move blank left, Right, Up, down
Goal Test
: Shown in Fig. Above
Path Cost
: 1 for each step
EA C461- Artificial Intelligence