FUZZY LOGIC

Fuzzy Logic
 Lotfi Zadeh (professor at UC Berkeley)
wrote his original paper on fuzzy set
theory. In various occasions, this is what
he said…
 “Fuzzy logic is a means of presenting
problems to computers in a way akin to the
way humans solve them”
 “The essence of fuzzy logic is that
everything is a matter of degree”
 What do these statements really mean?
Fuzzy Logic
 Very often, we humans analyze situations and solve
problems in a rather imprecise manner
 Do not have all the facts
 Facts might be uncertain

 Maybe we only generalize facts without having the

precise data or measurements…

 Real-life example: Playing a game of basketball

Everything is a matter of degree?
 Is your basketball opponent tall, or average or
short? (use of linguistic terms to measure degree)
 Is 7 feet tall? Is 6 feet 10 inches tall? Are they both
considered tall? (overlapping degrees)

 Problem with traditional Boolean logic

 You are forced to define a point above which we will
consider the guy to be tall or just average, e.g. > 7 ft
 Fuzzy Logic  allows gray areas or degrees of being
considered “tall”
The degree of truth
 So…you can think of fuzzy logic as classifying
something as being TRUE, but to varying degrees

 Real-life control applications (air-conditioning,

household appliances):
 Traditional Boolean logic will result in abrupt switching
of response functions
 Fuzzy logic alleviates this problem  Responses will
vary smoothly given the degree of truth or strength of
the input conditions
Fuzzy logic for games
 A previous game AI example…
 An AI character makes his decision to chase (using FSM
or DT) based on traditional Boolean logic, e.g. distance
of player < 20 units, and player health < 50%
 In fuzzy logic, we can represent these input conditions
using a few “membership” degrees of measure
 Distance: (“Far”, “Average”, “Near”)
 Health: (“Good”, “Normal”, “Poor”)

 The output actions can also be represented with

different membership degrees (“Chase Fast”, “Chase
Slow”)
How to use Fuzzy Logic in Games?
 3 possible ways how fuzzy logic can be used in
games
 Control
 Modulating steering forces, travelling/moving towards
target
 Threat Assessment
 Assessing
player’s strengths/weaknesses for deploying units
and making moves
 Classification
 Identifying
the combat prowess of characters in the game
based on a variety of factors in order to choose opponent
 There are many other possibilities…
Fuzzy Logic Basics
 Fuzzy control or fuzzy inference process – 3 basic
steps
Step 1: Fuzzification
 Fuzzification: Process of mapping/converting crisp
data (real numbers) to fuzzy data
 Find degree of membership of the crisp input in
predefined fuzzy sets
 E.g. given a character’s health, determine the degree to
which it is “Good”, “Fair” or “Poor”.
 Mapping is achieved using membership functions
Membership Functions
 Membership Functions
 Map input variables to a degree of membership, in a
fuzzy set, between 0 and 1. Degree 1  absolutely
true, degree 0  absolutely false, any degree in
between  true or false to a certain extent
 “Boolean logic membership function”
Membership Functions
 Fuzzy Membership Functions
 Enables us to transition gradually from false to true
 Grade membership function
Membership Functions
 Triangular m/f  Reverse grade m/f

 Equations are just the

inverse of the grade
m/f
Membership Functions
 Trapezoid m/f  Other nonlinear m/f
 Gaussian or Sigmoid
‘S’-shaped curves
Membership Functions
 Typically, we are interested in the degree of which
an input variable falls within a number of
qualitative sets
Membership Functions

 Setting up collections of fuzzy sets for an input variable

is a matter of judgment and trial-and-error  not
uncommon to “tune” the sets
 While tuning, one can try different membership
functions, increase or decrease number of sets
 Some fuzzy practitioners recommend 7 fuzzy sets to
fully define a practical working range (?!?!?)
Membership Functions

 One rule of thumb for ensuring smooth transitions (in

later steps) is to enforce overlapping between
neighboring sets
Hedge Functions
 Hedge functions are sometimes used to modify the
degree of membership
 Provide additional linguistic constructs that you can
use in conjunction with other logical operations.
 Two common hedges:
 VERY(Truth(A)) = Truth(A)2
 NOT_VERY(Truth(A)) = Truth(A)0.5

(Truth(A) is the degree of membership of A in some fuzzy

set)
Step 2: Fuzzy Rules
 Next, construct a set of rules, combining the input in
some logical manner, to yield some output
 If-then style rules (if A then B) – A being the
antecedent/premise and B being the
consequent/conclusion
 Fuzzy input variables are combined logically to
form premise
 Conclusion will be the degree of membership of
some output fuzzy set
Fuzzy Axioms
 Since we are writing “logical” rules with fuzzy input,
we need a way to apply logical operators to fuzzy
input (just like with Boolean input)
 Logical OR (disjunction)
 Truth(A OR B) = MAX(Truth(A), Truth(B))
 Logical AND (conjunction)
 Truth(A AND B) = MIN(Truth(A), Truth(B))
 Logical NOT (negation)
 Truth(NOT A) = 1 – Truth(A)
Fuzzy Axioms
 Example, given a person is overweight to the
degree of 0.7 and tall to the degree of 0.3:
 Overweight AND tall = MIN(0.7, 0.3) = 0.3
 Overweight OR tall = MAX(0.7, 0.3) = 0.7

 NOT overweight = 1 – 0.7 = 0.3

 NOT tall = 1 – 0.3 = 0.7

 NOT(overweight AND tall) = 1 – MIN(0.7, 0.3) = 0.7

 There are other definitions for these logical

operators…
Rule Evaluation
 Unlike traditional Boolean logic,
 Rules in fuzzy logic can evaluate into any number
between 0 and 1 (not just 0 or 1)
 All rules are evaluated in parallel (not in series that the
first one that is true gets fired). Each rule always fires,
to various degrees
 The strength of each rule represents the degree of
membership in the output fuzzy set
Rule Evaluation
 Example: Evaluating
whether an AI should
attack player
 Rules can be written like:
 If(in melee range AND
uninjured) AND NOT hard
then attack
 Set up as many rules to
handle all possibilities in
the game
Rule Evaluation
 Given specific degrees for the input variables, you
might get outputs (conclusions of the rules) that look
something like this:
Attack to degree: 0.2
Do nothing to degree: 0.4
Flee to degree: 0.7
 The most straightforward way to interpret these
outputs is to take the action associated with the
highest degree (in this case, the action will be flee)
Step 3: Defuzzification
 In some cases, you might want to use the fuzzy
output degree to determine a crisp value (real
number), which can be useful for further calculations
 Defuzzification: Process of converting the results
from the fuzzy rules to get a crisp number as an
output
 Opposite of fuzzification (you can say that,
although the purpose and methods are different!)
Step 3: Defuzzification
 Previous example: Instead of determining some
finite action (do nothing, flee, attack), we also want
to use the output to determine the speed to take the
action
 To get a crisp number, aggregate the output
strengths on the predefined output membership
functions
Step 3: Defuzzification

 With the numerical output from the earlier example

(0.2 degree attack, 0.4 degree do nothing, 0.7
degree flee), we have the composite membership
function below
Defuzzifying composite m/f

 Truncate each output set to the output degree of

membership for that set. Then combine all output sets
by disjunction
 A crisp number can be arrived from such an output
fuzzy set in many ways
 Geometric centroid of the area under the output fuzzy set,
taking its horizontal axis coordinate as the crisp output
Using “predefuzzified” output
 A less computationally expensive method is the use
of singleton output membership function or a
“predefuzzified” output function
 Instead of doing lots of calculation, assign speeds to
each output action (-10 for flee, 1 for do nothing,
10 for attack).
 E.g. The resulting speed for flee is simply the preset
value of -10 times the degree to which the output
action flee is true (-10 x 0.7 = -7)
Using “predefuzzified” output
 Aggregate of all outputs with a simple weighted
average
 In our example, we might have:
Output = [(0.7)(-10) + (0.4)(1) + (0.3)(10)] /
(0.7+0.4+0.3)
= -2.5
 This output would result in the creature fleeing, but
not earnestly in full extent
 Naturally, we can obtain various output (crisp)
values depending on the different input conditions
Further Examples
 There are 2 good examples in the textbook,
showing the full process of using fuzzy logic to
model game AI characters
Using Fuzzy Logic in FSMs?
 If we want to add some fuzzy logic into FSMs, how
can that be accomplish? Is it possible?
 Remember: Each state defines a behavior or action, and
each state is reached by transition from another state
on the basis of fulfilling some input conditions…
 Conditions for transition are normally in Boolean logic,
how do we accommodate fuzzy logic?
Fuzzy State Machines
 Different AI developers regard Fuzzy State
Machines differently
 State machine with fuzzy states
 State transitions that use fuzzy logic to trigger

 Both

 Find out more about how these different variations

can be worked out and implemented (refer to
Millington book)
…
 Next up
 Homework 2 (due in Week 10, submission via mail)
 Milestone #2 (due in Week 11, 11.00am 23/8, Thurs)

 Upcoming lectures
 Probabilitiesand Uncertainty Techniques
 Tactical and Strategic AI