Problem solving

CHAPTER
 Outline
2

Agents and environments

Rationality
PEAS (Performance measure, Environment,
Actuators, Sensors)
Environment types
Agent types

Artificial Intelligence a modern approach

 Problem solving
3

• A problem is defined by its elements and their

relations.
• Problem formulation is the process of deciding what
actions and states to consider, given a goal

Artificial Intelligence a modern approach

 Problem formulation
4

A problem can be defined formally by four

components:
1. Initial state
2. Successor function (Actions and Transition
model)
3. Goal test
4. Path cost

Artificial Intelligence a modern approach

 5

Initial State The starting state which agent knows

itself.
 Successor Function
 A description of the possible actions available to
the agent.
 State x, successor – FN (x) returns a set of
< action, successor> ordered pairs, where each
action is a legal action in a state x and each successor
is a state that can be reached from x by applying that
action.
Artificial Intelligence a modern approach
 State Space
6

The set of all possible states reachable from the

initial state by any sequence of actions.
The initial state and successor function
defines the state space.
The state space forms a graph in which nodes are
state and axis between the nodes are action.
Path A path in the state space is a sequence of state
connected by a sequence of actions.

Artificial Intelligence a modern approach

• State Space

A State space is the set of all states reachable from the

initial state. Definitions of terms :

◊ A state space forms a graph (or map) in which the nodes

are states and the arcs between nodes are actions.

◊ In state space, a path is a sequence of states connected by

a sequence of actions.

◊ The solution of a problem is part of the map formed by the

state space.
 Goal Test
8

Test to determine whether the given state is the goal

state.
If there is an explicit set of possible goal states, then
we can whether any one of the goal state is reached
or not.
Example : In chess, the goal is to reach a state called
―checkmateǁ where the opponent‘s king is under
attack and can‘t escape.

Artificial Intelligence a modern approach

 Path cost
9

 A function that assigns a numeric cost to each path.

 The cost of a path can be described as the sum of the costs
of the individual actions along that path.
 Step cost of taking an action ‗a‘ to go from one state ‗x‘ to
state ‗y‘ is denoted by C(x,a,y)
 C-Cost , x,y- states , Action , Step costs are non-negative
 These 4 elements are gathered into a data structure
that is given as input to problem solving algorithm.
 A solution quality is measured by path cost function. An
optimal solution has lowest path cost among all
solutions.

Artificial Intelligence a modern approach

 Problem types
10

 1. Single –state problem

 2. Multiple –state problem
 3. Contingency problem.
 Single –state problem
  observable (at least initial state)
  deterministic
  static
  discrete
 Multiple –state problem
  partially observable (initial state not observable)
  deterministic
 Static
  discrete Contingency problem  partially observable (initial state not
observable)  non-deterministic

Artificial Intelligence a modern approach

 11

 Search is fundamental to the problem-solving process.

  Search is a general mechanism that can be used when
more direct method is not known.
  Search provides the framework into which more direct
methods for solving subparts of a problem can be
embedded.
 A very large number of AI problems are formulated as
search problems.

Artificial Intelligence a modern approach

 Toy Problems
The vacuum world
12

 The vacuum world 1 2

 The world has only two
locations
 Each location may or may 3 4
not contain dirt
 The agent may be in one
location or the other 5 6
 8 possible world states
 Three possible actions:
Left, Right, Suck 7 8
 Goal: clean up all the dirt
 Toy Problems
The vacuum world
13

 States: one of the 8 states given earlier

 Operators: move left, move right, suck
 Goal test: no dirt left in any square
 Path cost: each action costs one

R
S S
L
R R
SS
L L

R
L
 8 puzzle problem
1. Examples of Problem
Definitions
 8-puzzle problem
15

Artificial Intelligence a modern approach

Artificial Intelligence a modern approach 16
Artificial Intelligence a modern approach 17
Artificial Intelligence a modern approach 18
 Problem Solving, Search
•
and Control Strategies
Exampl
e2:
 Search strategies
21

Searching is a process to find the solution for

a given set of problems.
 Rational agents or Problem-solving agents in
AI mostly used these search strategies or algorithms
to solve a specific problem and provide the best
result. Problem-solving agents are the goal-based
agents and use atomic representation

Artificial Intelligence a modern approach

 Search strategies
22

 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 ∞)

Artificial Intelligence a modern approach

 Searching Strategies
24

Uninformed/Blind Informed

 The uninformed search  Informed search algorithms

does not contain any use domain knowledge.
domain knowledge such as  In an informed search,
closeness, the location of problem information is
the goal available which can guide
 It operates in a brute-force the search.
way as it only includes  Informed search strategies
information about how to can find a solution more
traverse the tree and how to efficiently than an
identify leaf and goal nodes uninformed search strategy
Artificial Intelligence a modern approach
 Search Strategies
25

Uninformed/Blind Informed

 Uninformed search applies a  Informed search is also

way in which search tree is called a Heuristic search.
searched without any
information about the search  A heuristic is a way which
space like initial state operators might not always be
and test for the goal, so it is also guaranteed for best
called blind search. solutions but guaranteed
 It examines each node of the
to find a good solution in
tree until it achieves the goal
node. reasonable time.

Artificial Intelligence a modern approach

 Searching Strategies
26

Uninformed/Blind Informed

 Breadth-first search  Greedy Search

 Uniform cost search
 A* Search
 Depth-first search
 Iterative deepening depth-first
search
 Bidirectional Search

Artificial Intelligence a modern approach

 Uninformed/Blind Search
27

Breadth-First Search(BFS)
In breadth-first search, the tree or the graph is
traversed breadthwise, i.e. it starts from a node
called search key and then explores all the
neighbouring nodes of the search key at that depth-
first and then moves to the next level nodes. It is
implemented using the queue data structure that
works on the concept of first in first out (FIFO). It is
a complete algorithm as it returns a solution if a
solution exists

Artificial Intelligence a modern approach

 28

The time complexity for breadth-first search is bd where

b (branching factor) is the average number of child
nodes for any given node and d is depth.
The disadvantage of this algorithm is that it requires a
lot of memory space because it has to store each level of
nodes for the next one. It may also check duplicate
nodes.
Example: If the search starts from root node A to
reach goal node G, then it will traverse A-B-C-D-G. It
traverses level wise, i.e. explores the shallowest node
first.
Artificial Intelligence a modern approach
 29

Advantages:
BFS will provide a solution if any solution exists.
If there are more than one solutions for a given
problem, then BFS will provide the minimal solution
which requires the least number of steps.
Disadvantages:
It requires lots of memory since each level of the tree
must be saved into memory to expand the next level.
BFS needs lots of time if the solution is far away from
the root node.
Artificial Intelligence a modern approach
Artificial Intelligence a modern approach 30
Artificial Intelligence a modern approach 31
 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)
Artificial Intelligence a modern approach 33
Artificial Intelligence a modern approach 34
Artificial Intelligence a modern approach 35
Artificial Intelligence a modern approach 36
Artificial Intelligence a modern approach 37