Transcript CS 561a: Introduction to Artificial Intelligence

Knowledge and reasoning – second part
•
•
•
•
•
•
Knowledge representation
Logic and representation
Propositional (Boolean) logic
Normal forms
Inference in propositional logic
Wumpus world example
SE 420
1
Knowledge-Based Agent
• Agent that uses prior or acquired
knowledge to achieve its goals
Domain independent algorithms
ASK
Inference engine
TELL
Knowledge Base
Domain specific content
• Can make more efficient decisions
• Can make informed decisions
• Knowledge Base (KB): contains a set of
representations of facts about the Agent’s
environment
• Each representation is called a sentence
• Use some knowledge representation
language, to TELL it what to know e.g.,
(temperature 72F)
• ASK agent to query what to do
• Agent can use inference to deduce new
facts from TELLed facts
SE 420
2
Generic knowledge-based agent
1.
TELL KB what was perceived
Uses a KRL to insert new sentences, representations of facts, into KB
2.
ASK KB what to do.
Uses logical reasoning to examine actions and select best.
SE 420
3
Wumpus world example
SE 420
4
Wumpus world characterization
• Deterministic?
• Accessible?
• Static?
• Discrete?
• Episodic?
SE 420
5
Wumpus world characterization
• Deterministic?
Yes – outcome exactly specified.
• Accessible?
No – only local perception.
• Static?
Yes – Wumpus and pits do not move.
• Discrete?
Yes
• Episodic?
(Yes) – because static.
SE 420
6
Exploring a Wumpus world
A= Agent
B= Breeze
S= Smell
P= Pit
W= Wumpus
OK = Safe
V = Visited
G = Glitter
SE 420
7
Exploring a Wumpus world
A= Agent
B= Breeze
S= Smell
P= Pit
W= Wumpus
OK = Safe
V = Visited
G = Glitter
SE 420
8
Exploring a Wumpus world
A= Agent
B= Breeze
S= Smell
P= Pit
W= Wumpus
OK = Safe
V = Visited
G = Glitter
SE 420
9
Exploring a Wumpus world
A= Agent
B= Breeze
S= Smell
P= Pit
W= Wumpus
OK = Safe
V = Visited
G = Glitter
SE 420
10
Exploring a Wumpus world
A= Agent
B= Breeze
S= Smell
P= Pit
W= Wumpus
OK = Safe
V = Visited
G = Glitter
SE 420
11
Exploring a Wumpus world
A= Agent
B= Breeze
S= Smell
P= Pit
W= Wumpus
OK = Safe
V = Visited
G = Glitter
SE 420
12
Exploring a Wumpus world
A= Agent
B= Breeze
S= Smell
P= Pit
W= Wumpus
OK = Safe
V = Visited
G = Glitter
SE 420
13
Exploring a Wumpus world
A= Agent
B= Breeze
S= Smell
P= Pit
W= Wumpus
OK = Safe
V = Visited
G = Glitter
SE 420
14
Other tight spots
SE 420
15
Another example solution
No perception  1,2 and 2,1 OK
B in 2,1  2,2 or 3,1 P?
Move to 2,1
1,1 V  no P in 1,1
Move to 1,2 (only option)
SE 420
16
Example solution
S and No S when in 2,1  1,3 or 1,2 has W
1,2 OK  1,3 W
No B in 1,2  2,2 OK & 3,1 P
SE 420
17
Logic in general
SE 420
18
Types of logic
SE 420
19
The Semantic Wall
Physical Symbol System
World
+BLOCKA+
+BLOCKB+
+BLOCKC+
P1:(IS_ON +BLOCKA+ +BLOCKB+)
P2:((IS_RED +BLOCKA+)
SE 420
20
Truth depends on Interpretation
Representation 1
World
A
B
ON(A,B) T
ON(A,B) F
ON(A,B) F
A
ON(A,B) T
B
SE 420
21
Entailment
Entailment is different than inference
SE 420
22
Logic as a representation of the World
Representation: Sentences
entails
Sentence
Refers to
(Semantics)
World
Facts
SE 420
follows
Fact
23
Models
SE 420
24
Inference
SE 420
25
Basic symbols
• Expressions only evaluate to either “true” or “false.”
•
•
•
•
•
•
P
¬P
PVQ
P^Q
P => Q
PQ
“P is true”
“P is false”
negation
“either P is true or Q is true or both”
disjunction
“both P and Q are true”
conjunction
“if P is true, the Q is true”
implication
“P and Q are either both true or both false” equivalence
SE 420
26
Propositional logic: syntax
SE 420
27
Propositional logic: semantics
SE 420
28
Truth tables
• Truth value: whether a statement is true or false.
• Truth table: complete list of truth values for a statement given all
possible values of the individual atomic expressions.
Example:
P
T
T
F
F
Q
T
F
T
F
PVQ
T
T
T
F
SE 420
29
Truth tables for basic connectives
P Q
¬P
¬Q
PVQ
P ^ Q P=>Q PQ
T
T
F
F
F
F
T
T
F
T
F
T
T
T
T
F
T
F
F
F
T
F
T
F
SE 420
T
F
T
T
T
F
F
T
30
Propositional logic: basic manipulation rules
• ¬(¬A) = A
Double negation
• ¬(A ^ B) = (¬A) V (¬B)
• ¬(A V B) = (¬A) ^ (¬B)
Negated “and”
Negated “or”
•
•
•
•
•
•
Distributivity of ^ on V
by definition
using negated or
by definition
using negated and & or
A ^ (B V C) = (A ^ B) V (A ^ C)
A => B = (¬A) V B
¬(A => B) = A ^ (¬B)
A  B = (A => B) ^ (B => A)
¬(A  B) = (A ^ (¬B))V(B ^ (¬A))
…
SE 420
31
Propositional inference: enumeration method
SE 420
32
Enumeration: Solution
SE 420
33
Propositional inference: normal forms
“product of sums of
simple variables or
negated simple variables”
“sum of products of
simple variables or
negated simple variables”
SE 420
34
Deriving expressions from functions
• Given a boolean function in truth table form, find a propositional
logic expression for it that uses only V, ^ and ¬.
• Idea: We can easily do it by disjoining the “T” rows of the truth
table.
Example: XOR function
P
T
T
F
F
Q
T
F
T
F
RESULT
F
T
T
F
P ^ (¬Q)
(¬P) ^ Q
RESULT = (P ^ (¬Q)) V ((¬P) ^ Q)
SE 420
35
A more formal approach
• To construct a logical expression in disjunctive normal form from a
truth table:
- Build a “minterm” for each row of the table, where:
- For each variable whose value is T in that row, include
the variable in the minterm
- For each variable whose value is F in that row, include
the negation of the variable in the minterm
- Link variables in minterm by conjunctions
- The expression consists of the disjunction of all minterms.
SE 420
36
Example: adder with carry
Takes 3 variables in: x, y and ci (carry-in); yields 2 results: sum (s) and carryout (co). To get you used to other notations, here we assume T = 1, F =
0, V = OR, ^ = AND, ¬ = NOT.
co is:
s is:
SE 420
37
Tautologies
• Logical expressions that are always true. Can be simplified out.
Examples:
T
TVA
A V (¬A)
¬(A ^ (¬A))
AA
((P V Q)  P) V (¬P ^ Q)
(P  Q) => (P => Q)
SE 420
38
Validity and satisfiability
Theorem
SE 420
39
Proof methods
SE 420
40
Inference Rules
SE 420
41
Inference Rules
SE 420
42
Resolution
Conjunctive Normal Form (CNF)
•
conjunction of disjunctions of literals clauses
E.g., (A  B)  (B  C  D)
Resolution inference rule (for CNF):
li …  lk, m1  …  mn
li  …  li-1  li+1  …  lk  m1  …  mj-1  mj+1 ...  mn
where li and mj are complementary literals.
E.g., P1,3  P2,2, P2,2
P1,3
•
Resolution is sound and complete
for propositional logic
Conversion to CNF
B1,1  (P1,2  P2,1)
1)
Eliminate , replacing α  β with (α  β)(β  α).
2)
Eliminate , replacing α  β with α β.
3)
Move  inwards using de Morgan's rules and double-negation:
4)
Apply distributivity law ( over ) and flatten:
(B1,1  (P1,2  P2,1))  ((P1,2  P2,1)  B1,1)
(B1,1  P1,2  P2,1)  ((P1,2  P2,1)  B1,1)
(B1,1  P1,2  P2,1)  ((P1,2  P2,1)  B1,1)
(B1,1  P1,2  P2,1)  (P1,2  B1,1)  (P2,1  B1,1)
Resolution algorithm
• Proof by contradiction, i.e., show KBα unsatisfiable
Resolution example
• KB = (B1,1  (P1,2 P2,1))  B1,1
•
α = P1,2
Forward and backward chaining
• Horn Form (restricted)
KB = conjunction of Horn clauses
• Horn clause =
• proposition symbol; or
• (conjunction of symbols)  symbol
• E.g., C  (B  A)  (C  D  B)
•
• Modus Ponens (for Horn Form): complete for Horn KBs
α 1  …  αn  β
α1, … ,αn
β
• Can be used with forward chaining or backward chaining.
• These algorithms are very natural and run in linear time
Forward chaining
• Idea: fire any rule whose premises are satisfied in the
KB,
• add its conclusion to the KB, until query is found
Forward chaining algorithm
• Forward chaining is sound and complete for
Horn KB
Forward chaining example
Forward chaining example
Forward chaining example
Forward chaining example
Forward chaining example
Forward chaining example
Forward chaining example
Forward chaining example
Proof of completeness
•
FC derives every atomic sentence that is
entailed by KB
1. FC reaches a fixed point where no new atomic
sentences are derived
2. Consider the final state as a model m, assigning
true/false to symbols
3. Every clause in the original KB is true in m
a1  …  ak  b
4. Hence m is a model of KB
5. If KB╞ q, q is true in every model of KB, including
m
Backward chaining
Idea: work backwards from the query q:
to prove q by BC,
check if q is known already, or
prove by BC all premises of some rule concluding q
Avoid loops: check if new subgoal is already on the goal
stack
Avoid repeated work: check if new subgoal
1. has already been proved true, or
2. has already failed
Backward chaining example
Backward chaining example
Backward chaining example
Backward chaining example
Backward chaining example
Backward chaining example
Backward chaining example
Backward chaining example
Backward chaining example
Backward chaining example
Forward vs. backward chaining
• FC is data-driven, automatic, unconscious processing,
• e.g., object recognition, routine decisions
• May do lots of work that is irrelevant to the goal
• BC is goal-driven, appropriate for problem-solving,
• e.g., Where are my keys? How do I get into a PhD program?
• Complexity of BC can be much less than linear in size of
KB
Efficient propositional inference
Two families of efficient algorithms for propositional
inference:
Complete backtracking search algorithms
• DPLL algorithm (Davis, Putnam, Logemann, Loveland)
• Incomplete local search algorithms
• WalkSAT algorithm
The DPLL algorithm
Determine if an propositional logic sentence (in CNF) is satisfiable.
Improvements over truth table enumeration:
1. Early termination
A clause is true if any literal is true.
A sentence is false if any clause is false.
2. Pure symbol heuristic
Pure symbol: always appears with the same "sign" in all clauses.
e.g., In the three clauses (A  B), (B  C), (C  A), A and B are pure, C is
impure.
Make a pure symbol literal true.
3. Unit clause heuristic
Unit clause: only one literal in the clause
The only literal in a unit clause must be true.
The DPLL algorithm
The WalkSAT algorithm
• Incomplete, local search algorithm
• Evaluation function: The min-conflict heuristic of
minimizing the number of unsatisfied clauses
• Balance between greediness and randomness
The WalkSAT algorithm
Inference-based agents in the wumpus world
A wumpus-world agent using propositional logic:
P1,1
W1,1
Bx,y  (Px,y+1  Px,y-1  Px+1,y  Px-1,y)
Sx,y  (Wx,y+1  Wx,y-1  Wx+1,y  Wx-1,y)
W1,1  W1,2  …  W4,4
W1,1  W1,2
W1,1  W1,3
…
 64 distinct proposition symbols, 155 sentences
Wumpus world: example
• Facts: Percepts inject (TELL) facts into the KB
• [stench at 1,1 and 2,1]  S1,1 ; S2,1
• Rules: if square has no stench then neither the square
or adjacent square contain the wumpus
• R1: !S1,1 !W1,1  !W1,2  !W2,1
• R2: !S2,1 !W1,1 !W2,1  !W2,2  !W3,1
• …
• Inference:
• KB contains !S1,1 then using Modus Ponens we infer
!W1,1  !W1,2
 !W2,1
• Using And-Elimination we get: !W1,1
• …
SE 420
!W1,2
!W2,1
78
Limitations of Propositional Logic
1. It is too weak, i.e., has very limited expressiveness:
• Each rule has to be represented for each situation:
e.g., “don’t go forward if the wumpus is in front of you” takes 64
rules
2. It cannot keep track of changes:
• If one needs to track changes, e.g., where the agent has been
before then we need a timed-version of each rule. To track 100
steps we’ll then need 6400 rules for the previous example.
Its hard to write and maintain such a huge rule-base
Inference becomes intractable
SE 420
79
Summary
SE 420
80