Announcements
▪ Project 0 (optional) is due Tuesday, January 24, 11:59 PM PT
▪ HW0 (optional) is due Friday, January 27, 11:59 PM PT
▪ Project 1 is due Tuesday, January 31, 11:59 PM PT
▪ HW1 is due Friday, February 3, 11:59 PM PT
CS 188: Artificial Intelligence
Informed Search
Fall 2022
University of California, Berkeley
Today
▪ Informed Search
▪ Heuristics
▪ Greedy Search
▪ A* Search
▪ Graph Search
Recap: Search
Recap: Search
▪ Search problem:
▪ States (configurations of the world)
▪ Actions and costs
▪ Successor function (world dynamics)
▪ Start state and goal test
▪ Search tree:
▪ Nodes: represent plans for reaching states
▪ Plans have costs (sum of action costs)
▪ Search algorithm:
▪ Systematically builds a search tree
▪ Chooses an ordering of the fringe (unexplored nodes)
▪ Optimal: finds least-cost plans
Example: Pancake Problem
Cost: Number of pancakes flipped
Example: Pancake Problem
Example: Pancake Problem
State space graph with costs as weights
4
2 3
2
3
4
3
4 2
3 2
2
4
3
General Tree Search
Action: flip top two Action: fliptoallreach
Path four goal:
Cost: 2 Cost:
Flip 4 flip three
four,
Total cost: 7
Informed Search
Search Heuristics
▪ A heuristic is:
▪ A function that estimates how close a state is to a goal
▪ Designed for a particular search problem
▪ Examples: Manhattan distance, Euclidean distance for
pathing
10
5
11.2
Example: Heuristic Function
h(x)
Example: Heuristic Function
Heuristic: the number of the largest pancake that is still out of place
3
4
h(x)
3
4
3 0
4
4 3
4
4 2
3
Greedy Search
Greedy Search
▪ Expand the node that seems closest…
▪ What can go wrong?
Greedy Search
b
▪ Strategy: expand a node that you think is …
closest to a goal state
▪ Heuristic: estimate of distance to nearest goal for
each state
▪ A common case:
b
▪ Best-first takes you straight to the (wrong) goal …
▪ Worst-case: like a badly-guided DFS
[Demo: contours greedy empty (L3D1)]
[Demo: contours greedy pacman small maze (L3D4)]
A* Search
A* Search
UCS Greedy
A*
Combining UCS and Greedy
▪ Uniform-cost orders by path cost, or backward cost g(n)
▪ Greedy orders by goal proximity, or forward cost h(n)
8 g=0
S h=6
h=1 g=1
e a
1 h=5
1 3 2 g=2 g=9
S a d G
h=6 b d g=4 e h=1
h=6 h=5 h=2
1 h=2 h=0
1 g=3 g=6
c b g = 10
h=7 c G h=0 d
h=2
h=7 h=6
g = 12
G h=0
▪ A* Search orders by the sum: f(n) = g(n) + h(n)
Example: Teg Grenager
When should A* terminate?
▪ Should we stop when we enqueue a goal?
h=2
2 A 2
S h=3 h=0 G
2 B 3
h=1
▪ No: only stop when we dequeue a goal
Is A* Optimal?
h=6
1 A 3
S h=7
G h=0
▪ What went wrong?
▪ Actual bad goal cost < estimated good goal cost
▪ We need estimates to be less than actual costs!
Admissible Heuristics
Idea: Admissibility
Inadmissible (pessimistic) heuristics break Admissible (optimistic) heuristics slow down
optimality by trapping good plans on the fringe bad plans but never outweigh true costs
Admissible Heuristics
▪ A heuristic h is admissible (optimistic) if:
where is the true cost to a nearest goal
▪ Examples:
4
15
▪ Coming up with admissible heuristics is most of what’s involved
in using A* in practice.
Optimality of A* Tree Search
Optimality of A* Tree Search
Assume:
…
▪ A is an optimal goal node
▪ B is a suboptimal goal node
▪ h is admissible
Claim:
▪ A will exit the fringe before B
Optimality of A* Tree Search: Blocking
Proof:
…
▪ Imagine B is on the fringe
▪ Some ancestor n of A is on the
fringe, too (maybe A!)
▪ Claim: n will be expanded before B
1. f(n) is less or equal to f(A)
Definition of f-cost
Admissibility of h
h = 0 at a goal
Optimality of A* Tree Search: Blocking
1. f(n) is less than or equal to f(A)
…
▪ Definition of f-cost says:
f(n) = g(n) + h(n) = (path cost to n) + (est. cost of n to A)
f(A) = g(A) + h(A) = (path cost to A) + (est. cost of A to A)
▪ The admissible heuristic must underestimate the true cost
h(A) = (est. cost of A to A) = 0
▪ So now, we have to compare:
f(n) = g(n) + h(n) = (path cost to n) + (est. cost of n to A)
f(A) = g(A) = (path cost to A)
▪ h(n) must be an underestimate of the true cost from n to A
(path cost to n) + (est. cost of n to A) ≤ (path cost to A)
g(n) + h(n) ≤ g(A)
f(n) ≤ f(A)
Optimality of A* Tree Search: Blocking
Proof:
…
▪ Imagine B is on the fringe
▪ Some ancestor n of A is on the
fringe, too (maybe A!)
▪ Claim: n will be expanded before B
1. f(n) is less or equal to f(A)
2. f(A) is less than f(B)
B is suboptimal
h = 0 at a goal
Optimality of A* Tree Search: Blocking
2. f(A) is less than f(B)
…
▪ We know that:
f(A) = g(A) + h(A) = (path cost to A) + (est. cost of A to A)
f(B) = g(B) + h(B) = (path cost to B) + (est. cost of B to B)
▪ The heuristic must underestimate the true cost:
h(A) = h(B) = 0
▪ So now, we have to compare:
f(A) = g(A) = (path cost to A)
f(B) = g(B) = (path cost to B)
▪ We assumed that B is suboptimal! So
(path cost to A) < (path cost to B)
g(A) < g(B)
f(A) < f(B)
Optimality of A* Tree Search: Blocking
Proof:
…
▪ Imagine B is on the fringe
▪ Some ancestor n of A is on the
fringe, too (maybe A!)
▪ Claim: n will be expanded before B
1. f(n) is less or equal to f(A)
2. f(A) is less than f(B)
3. n expands before B
▪ All ancestors of A expand before B
▪ A expands before B
▪ A* search is optimal
Properties of A*
Properties of A*
Uniform-Cost A*
b b
… …
UCS vs A* Contours
▪ Uniform-cost expands equally in all
“directions”
Start Goal
▪ A* expands mainly toward the goal,
but does hedge its bets to ensure
optimality Start Goal
[Demo: contours UCS / greedy / A* empty (L3D1)]
[Demo: contours A* pacman small maze (L3D5)]
Comparison
Greedy Uniform Cost A*
A* Applications
▪ Video games
▪ Pathing / routing problems
▪ Resource planning problems
▪ Robot motion planning
▪ Language analysis
▪ Machine translation
▪ Speech recognition
▪ …
[Demo: UCS / A* pacman tiny maze (L3D6,L3D7)]
[Demo: guess algorithm Empty Shallow/Deep (L3D8)]
Creating Heuristics
Creating Admissible Heuristics
▪ Most of the work in solving hard search problems optimally is in coming up
with admissible heuristics
▪ Often, admissible heuristics are solutions to relaxed problems, where new
actions are available
366
15
▪ Inadmissible heuristics are often useful too
Example: 8 Puzzle
Start State Actions Goal State
▪ What are the states?
▪ How many states?
▪ What are the actions?
▪ How many successors from the start state?
▪ What should the costs be?
8 Puzzle I
▪ Heuristic: Number of tiles misplaced
▪ Why is it admissible?
▪ h(start) = 8
▪ This is a relaxed-problem heuristic
Start State Goal State
Average nodes expanded
when the optimal path has…
…4 steps …8 steps …12 steps
UCS 112 6,300 3.6 x 106
TILES 13 39 227
Statistics from Andrew Moore
8 Puzzle II
▪ What if we had an easier 8-puzzle where
any tile could slide any direction at any
time, ignoring other tiles?
▪ Total Manhattan distance
Start State Goal State
▪ Why is it admissible?
Average nodes expanded
▪ h(start) = 3 + 1 + 2 + … = 18 when the optimal path has…
…4 steps …8 steps …12 steps
TILES 13 39 227
MANHATTAN 12 25 73
8 Puzzle III
▪ How about using the actual cost as a heuristic?
▪ Would it be admissible?
▪ Would we save on nodes expanded?
▪ What’s wrong with it?
▪ With A*: a trade-off between quality of estimate and work per node
▪ As heuristics get closer to the true cost, you will expand fewer nodes but usually
do more work per node to compute the heuristic itself
Semi-Lattice of Heuristics
Trivial Heuristics, Dominance
▪ Dominance: ha ≥ hc if
▪ Heuristics form a semi-lattice:
▪ Max of admissible heuristics is admissible
▪ Trivial heuristics
▪ Bottom of lattice is the zero heuristic (what
does this give us?)
▪ Top of lattice is the exact heuristic
Graph Search
Tree Search: Extra Work!
▪ Failure to detect repeated states can cause exponentially more work.
State Graph Search Tree
Graph Search
▪ In BFS, for example, we shouldn’t bother expanding the circled nodes (why?)
d e p
b c e h r q
a a h r p q f
p q f q c G
q c G a
a
Graph Search
▪ Idea: never expand a state twice
▪ How to implement:
▪ Tree search + set of expanded states (“closed set”)
▪ Expand the search tree node-by-node, but…
▪ Before expanding a node, check to make sure its state has never been
expanded before
▪ If not new, skip it, if new add to closed set
▪ Important: store the closed set as a set, not a list
▪ Can graph search wreck completeness? Why/why not?
▪ How about optimality?
A* Graph Search Gone Wrong?
State space graph Search tree
A S (0+2)
1
1
S h=4
C
h=1 A (1+4) B (1+1)
h=2 1
2
3 C (2+1) C (3+1)
B
h=1
G G (5+0) G (6+0)
h=0
Consistency of Heuristics
▪ Main idea: estimated heuristic costs ≤ actual costs
A ▪ Admissibility: heuristic cost ≤ actual cost to goal
1 h(A) ≤ actual cost from A to G
h=4 C h=1
▪ Consistency: heuristic “arc” cost ≤ actual cost for each arc
h=2
h(A) – h(C) ≤ cost(A to C)
3
▪ Consequences of consistency:
▪ The f value along a path never decreases
G h(A) ≤ cost(A to C) + h(C)
▪ A* graph search is optimal
Optimality of A* Graph Search
Optimality
▪ Tree search:
▪ A* is optimal if heuristic is admissible
▪ UCS is a special case (h = 0)
▪ Graph search:
▪ A* optimal if heuristic is consistent
▪ UCS optimal (h = 0 is consistent)
▪ Consistency implies admissibility
▪ In general, most natural admissible heuristics
tend to be consistent, especially if from
relaxed problems
A*: Summary
A*: Summary
▪ A* uses both backward costs and (estimates of) forward costs
▪ A* is optimal with admissible / consistent heuristics
▪ Heuristic design is key: often use relaxed problems
Search and Models
▪ Search operates over
models of the world
▪ The agent doesn’t
actually try all the plans
out in the real world!
▪ Planning is all “in
simulation”
▪ Your search is only as
good as your models…
Search Gone Wrong?
Search Gone Wrong?
Appendix: Search Pseudo-Code
Tree Search Pseudo-Code
Graph Search Pseudo-Code
