Scribd
Upload
157 views

4 - C Problem Solving Agents

The document discusses problem solving agents in artificial intelligence. Problem solving agents use atomic representations to formulate goals based on the current situation and performance measures. They then search for solutions to reach the goal state by considering sequences of actions. Common search strategies include breadth-first search, depth-first search, and uniform-cost search for uninformed searches, and greedy best-first search and A* search for informed searches. Example problems that can be solved this way include the 8-puzzle and vacuum world problems.

Uploaded by

Pratik Raj
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
157 views

4 - C Problem Solving Agents

The document discusses problem solving agents in artificial intelligence. Problem solving agents use atomic representations to formulate goals based on the current situation and performance measures. They then search for solutions to reach the goal state by considering sequences of actions. Common search strategies include breadth-first search, depth-first search, and uniform-cost search for uninformed searches, and greedy best-first search and A* search for informed searches. Example problems that can be solved this way include the 8-puzzle and vacuum world problems.

Uploaded by

Pratik Raj
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 17

Artificial Intelligence

Problem Solving Agent


Contents
 Problem Solving Agent
 Problem Definition
 Example Problems
 Searching For Solutions
 Problem-solving Performance Measure
 Search Strategies
Problem Solving Agent
 The simplest agents were the reflex agents - base their actions
on a direct mapping from states to actions.
 cannot operate well in environments for which this mapping would be
too large to store
 Goal-based agents consider future actions and the desirability
of their outcomes
 One kind of goal-based agent called a problem-solving agent
 Problem-solving agents use atomic representations
 states of the world are considered as wholes, with no internal structure
visible
 Goal-based agents that use factored or structured
representations are usually called planning agents
Problem Solving Agent
 First Step in Problem Solving: Goal formulation
 based on the current situation and the agent’s performance
measure
 Courses of action that don’t achieve the goal can be rejected without
further consideration
 Goals help organize behavior by limiting the objectives
 The agent’s task is to find out how to act, now and in the future, so
that it reaches a goal state
 Problem formulation is the process of deciding what actions
and states to consider, given a goal.
 if the environment is unknown,
an agent with several immediate options of unknown
value can decide what to do by first examining future
actions that eventually lead to states of known value.
Kind of Environment
 Observable or partially observable?
 Discrete or Continuous?
 Deterministic or Stochastic?
 Static or Dynamic?
 Episodic or Sequential?
 Multiple or Single Agent?
 Competitive or Cooperative Agent?
Problem Solving Agent
 The process of looking for a sequence of actions that reaches
the goal - search.
 A search algorithm takes a problem as input and returns a
solution in the form of an action sequence.
 Once a solution is found, the actions it recommends can be
carried out - the execution phase.

“formulate, search, execute” design for the agent


 Steps:
 It first formulates a goal and a problem
 searches for a sequence of actions that would solve the problem, and
 then executes the actions one at a time.
 When this is complete, it formulates another goal and starts over.
Problem Definition
A problem can be defined formally by five components:
 Initial state - the agent starts in

 Actions
 A description of the possible actions available to the agent
 Given a particular state s, ACTIONS(s) returns the set of actions that
can be executed in s.
 Transition model
 A description of what each action does - the transition model
 Together, the initial state, actions, and transition model implicitly
define the state space of the problem
 state space forms a directed network or graph
 Goal test
 determines whether a given state is a goal state.
 Path cost
 function that assigns a numeric cost to each path.
Problem Definition
 A solution to a problem is an action sequence that leads from
the initial state to a goal state.
 Solution quality is measured by the path cost function, and
 an optimal solution has the lowest path cost among all
solutions.
Example Problems - 8-puzzle

9
8-puzzle
 States: the location of each of the eight tiles and the blank in one of the
nine squares
 Initial state: Any state can be designated as the initial state.
 Actions: The simplest formulation defines the actions as movements of the
blank space Left, Right, Up, or Down.
 Transition model: Given a state and action, this returns the resulting state;
 for example, if apply Left to the start state, the resulting state has the 5 and the blank
switched.
 Goal test: checks whether the state matches the goal configuration
 Path cost: Each step costs 1, so the path cost is the number of steps in the
path
Example Problems - vacuum world

11
vacuum world
 States: The state is determined by both the agent location and the dirt
locations.
 Initial state: Any state can be designated as the initial state.
 Actions: each state has just three actions: Left, Right, and Suck.
 Transition model: The actions have their expected effects, except that
moving Left in the leftmost square, moving Right in the rightmost
square, and Sucking in a clean square have no effect.
 Goal test: This checks whether all the squares are clean.
 Path cost: Each step costs 1, so the path cost is the number of steps in
the path.

• 8-queens problem
• route-finding problem
Searching for Solutions
 A solution is an action sequence, so search algorithms work
by considering various possible action sequences.
 The possible action sequences starting at the initial state
form a search tree with the initial state at the root;
 the branches are actions and
 the nodes correspond to states in the state space of the problem.
 Steps in growing Search tree
 root node of the tree corresponds to the initial state
 to test whether this is a goal state
 expanding the current state - generating a new set of states
 The set of all leaf nodes available for expansion at any given point - the frontier

 This is the essence of search


Tree-search and Graph-search

Three common variants are the


• first-in, first-out or FIFO queue
• the last-in, first-out or LIFO queue
• the priority queue 14
Performance Measure
Evaluate an algorithm’s performance in four ways
 Completeness: Is the algorithm guaranteed to find a solution when
there is one?
 Optimality: Does the strategy find the optimal solution?
 Time complexity: How long does it take to find a solution?
 No. of nodes generated during the search
 Space complexity: How much memory is needed to perform the
search?
 No. of nodes stored during the search

Complexity is expressed in terms of three quantities:


 b, the branching factor or maximum number of successors of any node
 d, the depth of the shallowest goal node
 m, the maximum length of any path in the state space.
Search Strategies
 uninformed search algorithms or blind search
 algorithms that are given no information about the problem other than
its definition.
 generate successors and distinguish a goal state from a non-goal state
 all search strategies are distinguished by the order in which nodes are
expanded
 although can solve any solvable problem, not so efficiently.
 Breadth-first search, DFS, Uniform-cost search
 Informed search algorithms or heuristic search
 strategies that know whether one non-goal state is “more promising”
than another
 some guidance on where to look for solutions
 Greedy best-first search, A* search
The End…

17

You might also like