Upload
rahul-sharma
View
224
Download
0
Embed Size (px)
Citation preview
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
1/79
Topic: Fuzzy LogicKnowledge Acquisition andInterface Design
Group 3:Rahul Sharma
Sumar Loomba
Nisheeth Gupta
Pallavi SagneRitu Khushwaha
Prashanth R
Praveen Rathod
Rohan Dange
Rakesh Kumar 1
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
2/79
FUZZY LOGIC
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
3/79
WHAT IS FUZZY LOGIC?
Definition of fuzzy
Fuzzynot clear, distinct, or precise; blurred
Definition of fuzzy logic
A form of knowledge representation suitable for notions that
cannot be defined precisely, but which depend upon their
contexts.
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
4/79
TRADITIONAL REPRESENTATION OF
LOGIC
Slow Fast
Speed = 0 Speed = 1bool speed;
get the speed
if ( speed == 0) {
// speed is slow}
else {
// speed is fast
}
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
5/79
FUZZY LOGIC REPRESENTATION
For every problem
must represent in terms
of fuzzy sets.
What are fuzzy sets?
Slowest
Fastest
Slow
Fast
[ 0.00.25
]
[ 0.250.50 ]
[ 0.500.75 ]
[ 0.751.00 ]
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
6/79
FUZZY LOGIC REPRESENTATION CONT.
Slowest Fastestfloat speed;
get the speed
if ((speed >= 0.0)&&(speed < 0.25)) {
// speed is slowest
}
else if ((speed >= 0.25)&&(speed < 0.5))
{
// speed is slow
}else if ((speed >= 0.5)&&(speed < 0.75))
{
// speed is fast
}
else // speed >= 0.75 && speed < 1.0
{
// speed is fastest
}
Slow Fast
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
7/79
ORIGINS OF FUZZY LOGIC
Traces back to Ancient Greece
Lotfi Asker Zadeh ( 1965 )
First to publish ideas of fuzzy logic.
Professor Toshire Terano ( 1972 )
Organized the world's first working group on fuzzy systems.
F.L. Smidth & Co. ( 1980 ) First to market fuzzy expert systems.
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
8/79
FUZZY LOGIC IN CONTROL SYSTEMS
Fuzzy Logic provides a more efficient and resourceful
way to solve Control Systems.
Some Examples
Temperature Controller
AntiLock Break System ( ABS )
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
9/79
TEMPERATURE CONTROLLER
The problem
Change the speed of a heater fan, based off the room
temperature and humidity.
A temperature control system has four settings
Cold, Cool, Warm, and Hot
Humidity can be defined by:
Low, Medium, and High
Using this we can define
the fuzzy set.
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
10/79
BENEFITS OF USING FUZZY LOGIC
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
11/79
ANTI LOCK BREAK SYSTEM ( ABS )
Nonlinear and dynamic in nature
Inputs for Intel Fuzzy ABS are derived from
Brake
4 WD
Feedback
Wheel speed
Ignition
Outputs
Pulsewidth
Error lamp
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
12/79
FUZZY LOGIC IN OTHER FIELDS
Business
Hybrid Modeling
Expert Systems
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
13/79
13
Fuzzy Logic and Knowledge
Based Systems (AI)
Knowledge Acquisition
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
14/79
What is Knowledge Acquisition?
Knowledge acquisition(KA) is the process of acquiringknowledge from a human expert for an expert system whichmust be carefully organized into if-then else rules or some otherform of knowledge representation. KA is the process of
absorbing and storing new information in memory, the successof which depends on how well the information can later beretrieved from memory. The process of storing and retrievinginformation depends heavily on the representation andorganization of the information.
27/02/2013 14
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
15/79
15
Knowledge Acquisition INTRODUCTION& BACKGROUND
The important characteristics of knowledgeare that it isexperiential, descriptive, qualitative, largely undocumentedandconstantly changing.
There are certain domains where all these properties are found
and some where there are only a few.
The lack of documentation and the fact that experts carry a lot ofinformation in their heads, make it difficult to gain access to theirknowledge for developing information systems in general and
expert systems in particular.
Therefore, knowledge engineers have devised specialisedtechniques to extract and document this information in an efficient
and expedient manner: Knowledge Acquisition.
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
16/79
16
Knowledge AcquisitionINTRODUCTION& BACKGROUND
Currently knowledge bases for knowledge based systems are crafted
by hand, this is a severe limitation on the rapid deployment of such
systems.
The automation of knowledge acquisition (from text) would greatly ease
this problem.
There is considerable interest in developing software tools which would
allow the automatic construction of knowledge bases from textualinformation.
This will provide the opportunity to rapidly build knowledge bases thus
increasing, for example, the rate at which knowledge based systems can
be developed and deployed
Knowledge acquisition can be regarded as a method by which aknowledge engineer obtains information from experts, text books,
and other authoritative sources for ultimate translation into a
machine language and knowledge base.
The person undertaking the knowledge acquisition must convert the
acquired knowledge into a form that a computer program can use.
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
17/79
17
Knowledge AcquisitionINTRODUCTION& BACKGROUND
In the process of Knowledge Acquisition for an Expert System Project,
the knowledge engineer basically performs four major tasks in
sequence:
First, the engineer ensures that he or she understands the aims and objectives of
the proposed expert system to get a feeling for the potential scope of the
project.Second, he or she develops a working knowledge of the problem domain by
mastering it's terminology by looking up technical dictionaries and
terminology data bases. For this task the key sources of knowledge are
identified: textbooks, papers, technical reports, manuals, codes of practice,
and domain experts.
Third, the knowledge engineer interacts with experts via meetings or interviewsto acquire, verify and validate their knowledge.
Fourth, the knowledge engineer produces a "paper knowledge base"; a document
or group of documents which form an intermediate stage in the translation of
knowledge from source to computer program. This comprises the interview
transcripts, the analysis of the information they contain and a full descriptionof the major domain entities (e.g. tasks, rules and objects).
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
18/79
18
Knowledge AcquisitionINTRODUCTION& BACKGROUND
Knowledge engineers interview experts in a specialist domain about how
he or she solves a given problem. Before interviewing the experts, the
knowledge engineers have to formulate their questions, and after the
interview the answers to the questions have to be analyzed.
The knowledge engineer has to familiarize himself or herself with the
terminology of the specialist domain; he or she has to consult technical
manuals, and in some cases learned papers, to see how the experts knowledge
is applied; the knowledge engineers sometimes consults textbooks or
encyclopedic texts for understanding the conceptual structure of the experts
domain.
In many different ways the knowledge engineer literally has to come to
terms with the language used by the expert and that used in the other
texts mentioned above. The knowledge engineer should become
conversant in the specialist language of his or her application domain.
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
19/79
19
Knowledge acquisition involveselicitation, analysis, modelling andvalidation of knowledge
1. Employing a technique to elicit data (usually verbal) from the expert.
2. Interpreting these verbal data (more or less skilfully) in order to inferwhat might be the expert's underlying knowledge and reasoning
process.
3. Using this interpretation to guide the construction of some model orlanguage that describes (more or less accurately) the expert'sknowledge and performance.
4. Interpretation of further data is guided in turn by this evolvingmodel.
5. The principle focus for the knowledge acquisition team should be inconstructing models, in domain definition, or problem identificationand problem analysis.
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
20/79
20
Roles for knowledge acquisition
Knowledge engineering and management: technological
innovation, ontology construction, document mark-up
AI systems development: generic methodologies (e.g., KADS: KADSstands for ``Knowledge Analysis and Documentation System''. Later on,other interpretations have been given to this acronym, such as``Knowledge Analysis and Design Support'. KADS is the name of a
structured methodology for the development of knowledge based systemsthat is now in practical use in many places in Europe and elsewhere.)
Organizational analysis:process approaches
Task analysis:job design
User analysis: generation of cognitive specifications for tasks, themitigation of human error in domains of risk or time pressure, theenhancement of proficiency through training and skill remediation
Requirements elicitation: systems or design analysis, conceptualdatabase design, software requirements definition
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
21/79
21
Preliminary Work I involves Reading,Observation, Discussion
Preliminary work is carried out by knowledge engineer(s)
Knowledge engineering is knowledge acquisition for expert systemdevelopment, and used to describe the reduction of a large body ofknowledge to a precise set of facts and rules
Knowledge engineer is a computer software programmer who gathersknowledge from experts and then translates the knowledge into the
knowledge base of a computerised expert system in a structured and logicalway, and eventually constructs computerised expert systems.
"Knowledge acquisition is a bottleneck in the construction of expertsystems.The knowledge engineer's job is to act as a go-between to help anexpert build a system. Since the knowledge engineer has far less knowledge
of the domain than the expert, however, communication problems impedethe process of transferring expertise into a program. The vocabularyinitially used by the expert to talk about the domain with a novice is ofteninadequate for problem-solving; thus the knowledge engineer and expertmust work together to extend and refine it. One of the most difficult aspectsof the knowledge engineer's task is helping the expert to structure the
domain knowledge, to identify and formalize the domain concepts."(Ref:
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
22/79
22
Preliminary Work - II
When acquiring knowledge about a domain it is absolutely crucial that the
knowledge engineer can converse with the expert using the expertterminology.
The knowledge engineer has to have a good grasp of the domain to be ableto ask intelligent questions to extract important and relevant knowledgefrom the experts who have vast amounts of knowledge a lot of which istacit knowledge.
The knowledge engineer must therefore do some preliminary workincluding research on the domain in question before the first interview withthe expert takes place.
Some requirements for KA Techniques
Take experts off the job for short time periods
Allow non-experts to understand the knowledge Focus on the essential knowledge
Try to capture tacit knowledge
Allow knowledge to be collated from different experts
Allow knowledge to be validated and maintained
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
23/79
23
The Knowledge Handbook
One of the functions of the knowledge engineer during the knowledgeacquisition phase is to document the knowledge that has been acquired. One
idea suggested (Wolfgram et. al. 1987 and others) is that of building aknowledge handbook.
Wolfgram et. al. describe the contents of the knowledge handbook as follows:
The general problem description.
Who the users are and their expectations from the system.
A breakdown of the problems into sub-problems and sub-domains forfuture knowledge acquisition.
A detailed description of the domain or sub-domain to be used for the
prototype.
A bibliography of reference documents.
A list of vocabulary, concepts, terms, phrases and acronyms in the domain.
A list of experts for the prototype.
Some reasonable performance standards for the system, based on
consultation with the experts and users.
Descriptions of typical reasoning scenarios gained from the knowledgeacquisition.
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
24/79
24
Basic knowledge engineering forknowledge acquisition - I
Knowledge engineer act as a go-between the expert and knowledge base.This can be achieved by means of eliciting knowledge from the expert,encoding it for the knowledge base, and refining it in collaboration with theexpert in order to achieve acceptable performance. The process is basicallyas follows:
The knowledge engineer interviews the expert to elicit his or herknowledge;
The knowledge engineer encodes the elicited knowledge for theknowledge base;
The shell uses the knowledge base to make inferences about particularcases specified by clients;
The clients use the shell's inferences to obtain advice about particularcases
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
25/79
25
Basic knowledge engineering forknowledge acquisition - II
Basic knowledge engineering model with manual acquisition of knowledgefrom an expert (left-hand side of the figure). This is also followed byinteractive application of the knowledge with multiple clients through anexpert system shell (right-hand side of the figure).
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
26/79
26
Interactive Knowledge Engineering forInteractive Knowledge Acquisition -I
In an interactive knowledge engineering process for interactive knowledge
acquisition, knowledge engineers have responsibility for: Advising the experts on the process of interactive knowledge
elicitation;
Managing the interactive knowledge acquisition tools, setting them upappropriately;
Editing the uuencoded knowledge base in collaboration with theexperts;
Managing the knowledge encoding tools, setting them upappropriately;
Editing the encoded knowledge base in collaboration with the experts;
Validating the application of the knowledge base in collaboration with
the experts; Setting up the user interface in collaboration with the experts and
clients;
Training the clients in the effective use of the knowledge base incollaboration with the expert by developing operational and training
procedures.
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
27/79
27
Interactive Knowledge Engineering forInteractive knowledge acquisition -II
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
28/79
28
Interactive Knowledge Engineering forInteractive Knowledge Acquisition -IIIInteractive knowledge acquisition and encoding tools can greatly reduce the
need for the knowledge engineer to act as an intermediary but, in mostapplications, they leave a substantial role for the knowledge engineer.
This use of interactive elicitation can be combined with manual elicitation and
with the use of the interactive tools by knowledge engineers rather than, or in
addition to, experts. Knowledge engineers can directly elicit knowledge from
the expert and use the interactive elicitation tools to enter knowledge into the
knowledge base.
Such approach is very useful and effective as it allows use of
Multiple knowledge engineers since the tasks may require the effort of
more than one person, and some specialization may be appropriate
Multiple experts since one person (expert) should not be expected to
have all the knowledge required, and, even if such an expert exists,comparative elicitation from multiple experts is itself a valuable
knowledge elicitation technique
Validation process, which is a key to an effective and successful system
development
Knowledge Acquisition
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
29/79
29
Knowledge AcquisitionTasks performed by a knowledge
engineer
Domain Terminology
Salient domain features
ReviseLearn
Scope of the problem
Knowledge Sources
Outline Constrain
Problem-solving tasks
Domain objectsSpecify Verify
Paper Know ledge Base
Rules and HeuristicsProduce Validate
Overview Interview
FocussedInterview s
Literature Review
StructuredInterview
Rule Animation
Objectives Revision PhaseDiscovery Phase Technique Used
Consult Textbooks
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
30/79
Traditional Approach to Knowledge Acquisition
In the traditional approach to acquiring knowledge, aknowledge engineer consults reference materials, databases,and human experts.
The knowledge captured will be both explicit and tacitknowledge.
Explicit knowledge is acquired through printed material.
Tacit knowledge originates from human resources. It is thetacit knowledge that never gets quantified into a manual orother accessible form, but resides in the minds of the peoplewho have worked with and developed that information.
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
31/79
Tacit Knowledge Acquisition
Traditional Methods for Tacit knowledge acquisition are Costlybecause at least two (typically) expensive people are involved, i.e.,the domain expert and the knowledge engineer.
The methods are error prone because people cant easily say what it
is that they do in a manner that can be understood by others
Traditional Methods is time-consuming because errors, gaps, andinconsistencies may be difficult to discover, requiring manyinteractions between experts and knowledge engineers to debug afield-ready application.
Clearly, costs must be reduced, errors eliminated, and developmenttime shortened. An approach to solve these issues is to augment theknowledge engineer with a framework for knowledge acquisition.
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
32/79
Explicit Knowledge Acquisition
There are often several sources of explicit knowledge
1. Literature : These documents can also be helpful in defining and clarifying theterminology of the problem domain.
2. Company Policy Manuals and Regulations : These documents are generallyformatted and organized in a manner that is analogous to the format and organizationof business rules specifications.
3. Reports, Memos and Guidelines : These types of documents are generally notformatted and organized in a manner that is useful to the elicitation process.
4. Published Books and Journal Articles : Published sources are generally the leastuseful forms of documentation to the elicitation process.
5. Existing Application Code
6. Database-Stored Procedures
7. Program Source Code
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
33/79
Tools for KnowledgeAcquisition
-Rahul sharma
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
34/79
Elicitation Methods
Manual Based on interview Track reasoning process Observation
Semiautomatic Build base with minimal help from
knowledge engineerAllows execution of routine tasks with
minimal expert input
Automatic
Minimal input from both expert andknowled e en ineer
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
35/79
Manual Methods
Interviews
Structured
Goal-orientedWalk through
Unstructured
Complex domains
Data unrelated and difficult to integrate
Semistructured
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
36/79
Manual Methods
Process tracking
Track reasoning processes
Protocol analysis Document experts decision-making
Think aloud process
Observation Motor movements
Eye movements
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
37/79
Manual Methods
Case analysis
Critical incident
User discussions
Expert commentary
Graphs and conceptual modelsBrainstorming
Prototyping
Multidimensional scaling for distance matrix
Clustering of elements
Iterative performance review
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
38/79
Semiautomatic Methods
Repertory grid analysis Personal construct theory
Organized, perceptual model of experts knowledge Expert identifies domain objects and their attributes Expert determines characteristics and opposites for each
attribute Expert distinguishes between objects, creating a grid
Expert transfer system
Computer program that elicits information fromexperts Rapid prototyping Used to determine sufficiency of available
knowledge
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
39/79
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
40/79
Semiautomatic Methods, continued
Computer based tools features:
Ability to add knowledge to base
Ability to assess, refine knowledgeVisual modeling for construction of domain
Creation of decision trees and rules
Ability to analyze information flows Integration tools
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
41/79
Automatic Methods
Data mining by computers
Inductive learning from existing
recognized casesNeural computing mimicking humanbrain
Genetic algorithms using naturalselection
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
42/79
Neural network (MLP)
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
43/79
Model of an artificial neuron
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
44/79
General neuro-fuzzy architecture
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
45/79
Multiple Experts
Scenarios Experts contribute individually
Primary experts information reviewed by
secondary experts Small group decision
Panels for verification and validation
Approaches Consensus methods Analytic approaches
Automation of process through software usage
Decomposition
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
46/79
Automated Knowledge Acquisition
Induction
Activities
Training set with known outcomes Creates rules for examples
Assesses new cases
Advantages
Limited application
Builder can be expert
Saves time, money
Automated Knowledge
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
47/79
Automated KnowledgeAcquisition
Difficulties Rules may be difficult to understand
Experts needed to select attributes
Algorithm-based search process producesfewer questions
Rule-based classification problems
Allows few attributes
Many examples needed Examples must be cleansed
Limited to certainties
Examples may be insufficient
Automated Knowledge
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
48/79
Automated KnowledgeAcquisition
Interactive induction
Incrementally induced knowledge
General models Object Network
Based on interaction with expert
interviews
Computer supported Induction tables
IF-THEN-ELSE rules
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
49/79
Practical Application ofFuzzy Logic
Rakesh kumar,108
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
50/79
Fault Analysis
The fuzzy detection system is developed and tested withnoisy data and with filtered data. It detects on filtereddata i.e. very accurate with no false alarms andnegligible missed alarms.
Decision Making
A new decision making method using fuzzy logic isproposed. The objective is to solve behaviour conflicts in
behaviour-based architectures. Two main problems havebeen identified: how to decide which behaviour shouldbe activated at each instant; and how to combine the
results from different behaviours into one action.
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
51/79
Image Analysis
fuzzy technique, was chosen for imageenhancement and applied in a very specific fieldof optical measurements
SchedulingThe scheduling and mapping of the precedence-constrained task graphs of parallel programs toprocessors is considered one of the most crucial NP-
complete problems in parallel and distributed computingsystems. task scheduling model based on fuzzy logic isproposed
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
52/79
. Problem Solving MethodologyIt is used directly and indirectly in a no. of applications
Fraud Detection
The system detects probable fraudulent behaviour:
by evaluating all the characteristics of a provider's claim data inparallel,
against the normal behaviour of a small( in demographic terms )community.
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
53/79
Soft Computing
Soft computing plays an increasingly important role inmany application areas, including software engineering.
Intelligent RoboticsThe use of fuzzy logic has resulted in smooth motion
control robust performance in face of errors in the priorknowledge and in the sensor data and principledintegration between different layers.
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
54/79
Speech Recognition
To apply fuzzy logic to speech recognition is a new attempt in digitalspeech processing. The approach proposed in the paper simplifies thealgorithm in speech recognition and makes the real-time processingtime shorter.
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
55/79
Practical Applications
Captures intuitive, human expressions
Precision temperature and humidity control
Straightforward and therefore inexpensiveControl electronics
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
56/79
Practical Applications
Electric Power
Industrial applications
Automated Cow Status Monitoring
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
57/79
Practical Applications
Applied to motor control
Minimization of cycling times
Fuzzy logic Design
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
58/79
Knowledge Representation
Pallavi Sagne , 92
Ritu Kushwaha,112
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
59/79
Fuzzy logic is a set of mathematical principles forknowledge representation based on degrees ofmembership.
Unlike two-valued Boolean logic, fuzzy logic is multi-valued. It deals with degrees of membership and degrees of
truth. Logical values between 0 (completely false) and 1
(completely true). Accepting that things can be partly true and partly
false at the same time.
Range of logical values in
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
60/79
Range of logical values inBoolean and fuzzy logic
(a) Boolean Logic. (b) Multi-valued Logic.
0 1 10 0.2 0.4 0.6 0.8 100 1 10
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
61/79
The classical example in fuzzy sets is tall men.The elements of the fuzzy set tall men are all men, but theirdegrees of membership depend on their height.
Degree of Membership
Fuzzy
Mark
John
Tom
Bob
Bill
1
1
1
0
0
1.00
1.00
0.98
0.82
0.78
Peter
Steven
Mike
David
Chris
Crisp
1
0
0
0
0
0.24
0.15
0.06
0.01
0.00
ame Height, cm
205
198
181
167
155
152
158
172
179
208
Crisp and fuzzy sets of tall
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
62/79
C sp a d u y se s omen
150 210170 180 190 200160
Height, cmDegree of
embership
Tall Men
150 210180 190 200
1.0
0.0
0.2
0.4
0.6
0.8
160
Degree ofembership
170
1.0
0.0
0.2
0.4
0.6
0.8
Height, cm
Fuzzy Sets
Crisp Sets
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
63/79
The x-axis represents the universe of discourse
The range of all possible values applicable to a chosenvariable.
In our case, the variable is the man height.
According to this representation, the universe of mensheights consists of all tall men.
The y-axis represents the membership value of the fuzzy
set. In our case, the fuzzy set of tall men maps height
values into corresponding membership values.
A fuzzy set is a set with fuzzy boundaries
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
64/79
A fuzzy set is a set with fuzzy boundaries.
Let X be the universe of discourse and its elements be
denoted as x.
In the classical set theory, crisp set A of X is defined asfunction fA(x) called the characteristic function of A
fA(x): X {0, 1}, where
Ax
AxxfA
if0,
if1,)(
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
65/79
In the fuzzy theory, fuzzy set A of universe X is defined byfunction A(x) called the membership function of set A
A(x): X [0, 1], where A(x) = 1 if x is totally in A;
A(x) = 0 if x is not inA;
0 < A(x) < 1 if x is partly inA.
membership function A(x) equals the degree to which x is anelement of set A.
This degree, a value between 0 and 1, represents the degreeof membership, also called membership value, of element x inset A.
Crisp and fuzzy sets of short, average and
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
66/79
tall men
150 210170 180 190 200160
Height, cmDegree of
embership
Tall Men
150 210180 190 200
1.0
0.0
0.2
0.4
0.6
0.8
160
Degree ofembership
Short Average ShortTall
170
1.0
0.0
0.2
0.4
0.6
0.8
Fuzzy Sets
Crisp Sets
Short Average
Tall
Tall
Representation of crisp and fuzzy
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
67/79
p p ysubsets
Fuzzy Subset A
uzziness
1
0Cris Subset A uzziness
(x)
Typical functions that can be used to represent a fuzzy set
Sigmoid, Gaussian and pi.
These functions increase the time of computation.
In practice, most applications use linear fit functions.
Abd i
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
68/79
68
AbductionAbduction is a reasoning process that tries to form
plausible explanations for abnormal observations Abduction is distinctly different from deduction and induction
Abduction is inherently uncertain
Uncertainty is an important issue in abductivereasoning
Some major formalisms for representing and reasoningabout uncertainty Mycins certainty factors (an early representative)
Probability theory (esp. Bayesian belief networks)
Dempster-Shafer theory
Fuzzy logic Truth maintenance systems
Nonmonotonic reasoning
Abd ti
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
69/79
69
Abduction
Definition (Encyclopedia Britannica):reasoning that derives an explanatoryhypothesis from a given set of facts The inference result is a hypothesisthat, if true, could
explain the occurrence of the given facts
Examples Dendral, an expert system to construct 3D structure of
chemical compounds
Fact: mass spectrometer data of the compound and its
chemical formula KB: chemistry, esp. strength of different types of bounds
Reasoning: form a hypothetical 3D structure that satisfiesthe chemical formula, and that would most likely producethe given mass spectrum
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
70/79
70
Medical diagnosis
Facts: symptoms, lab test results, and other observedfindings (called manifestations)
KB: causal associations between diseases andmanifestations
Reasoning: one or more diseases whose presencewould causally explain the occurrence of the givenmanifestations
Many other reasoning processes (e.g., word sensedisambiguation in natural language process, imageunderstanding, criminal investigation) can also beenseen as abductive reasoning
Abduction examples (cont.)
Comparing abduction, deduction,
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
71/79
71
and induction
Deduction: major premise: All balls in the box are blackminor premise: These balls are from the box
conclusion: These balls are black
Abduction: rule: All balls in the box are black
observation: These balls are black
explanation: These balls are from the box
Induction: case: These balls are from the box
observation: These balls are black
hypothesized rule: All ball in the box are black
A => B
A---------B
A => BB
-------------
Possibly A
WheneverA then B-------------PossiblyA => B
Deductionreasons from causes to effectsAbduction reasons from effects to causesInduction reasons from specific cases to general rules
Characteristics of abductive
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
72/79
72
reasoning
Conclusions are hypotheses, not theorems (maybe false even ifrules and facts are true) E.g., misdiagnosis in medicine
There may be multiple plausible hypotheses Given rules A => B and C => B, and fact B, both A
and C are plausible hypotheses
Abduction is inherently uncertain Hypotheses can be ranked by their plausibility (if it
can be determined)
Characteristics of abductive
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
73/79
73
Characteristics of abductivereasoning (cont.)
Reasoning is often a hypothesize-and-test cycle Hypothesize: Postulate possible hypotheses, any of which
would explain the given facts (or at least most of theimportant facts)
Test: Test the plausibility of all or some of these hypotheses One way to test a hypothesis H is to ask whether something
that is currently unknownbut can be predicted from Hisactually true If we also know A => D and C => E, then ask if D and E are true If D is true and E is false, then hypothesis A becomes more plausible
(support for A is increased; support for C is decreased)
Characteristics of abductive
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
74/79
74
C a acte st cs o abduct ereasoning (cont.)
Reasoning is non-monotonic That is, the plausibility of hypotheses can
increase/decrease as new facts are collected
In contrast, deductive inference is monotonic: itnever change a sentences truth value, onceknown
In abductive (and inductive) reasoning, somehypotheses may be discarded, and new onesformed, when new observations are made
Sources of uncertainty
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
75/79
75
Sources of uncertainty
Uncertain inputs Missing data Noisy data
Uncertain knowledge Multiple causes lead to multiple effects Incomplete enumeration of conditions or effects Incomplete knowledge of causality in the domain Probabilistic/stochastic effects
Uncertain outputs Abduction and induction are inherently uncertain Default reasoning, even in deductive fashion, is uncertain Incomplete deductive inference may be uncertain
Probabilistic reasoning only gives probabilistic results(summarizes uncertainty from various sources)
Decision making with
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
76/79
76
guncertainty
Rational behavior: For each possible action, identify the possible
outcomes
Compute the probability of each outcome
Compute the utility of each outcome
Compute the probability-weighted (expected)utility over possible outcomes for each action
Select the action with the highest expected utility(principle ofMaximum Expected Utility)
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
77/79
77
Bayesian reasoning
Probability theory
Bayesian inference
Use probability theory and information about
independence Reason diagnostically (from evidence (effects) to
conclusions (causes)) or causally (from causes toeffects)
Bayesian networks Compact representation of probability distribution
over a set of propositional random variables
Take advantage of independence relationships
Other uncertainty representations
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
78/79
78
Other uncertainty representationsDefault reasoning Nonmonotonic logic: Allow the retraction of default beliefs if
they prove to be false
Rule-based methods Certainty factors (Mycin): propagate simple models of belief
through causal or diagnostic rules
Evidential reasoning Dempster-Shafer theory: Bel(P) is a measure of the evidence
for P; Bel(P) is a measure of the evidence against P;together they define a belief interval (lower and upperbounds on confidence)
Fuzzy reasoning Fuzzy sets: How well does an object satisfy a vague
property?
Fuzzy logic: How true is a logical statement?
Uncertainty tradeoffs
7/29/2019 Fuzzy Logic Knowledge Acquisition Final
79/79
Uncertainty tradeoffs
Bayesian networks: Nice theoretical properties
combined with efficient reasoning make BNs verypopular; limited expressiveness, knowledgeengineering challenges may limit uses
Nonmonotonic logic: Represent commonsense
reasoning, but can be computationally very expensiveCertainty factors: Not semantically well founded
Dempster-Shafer theory: Has nice formalproperties, but can be computationally expensive,and intervals tend to grow towards [0,1] (not a very
useful conclusion)Fuzzy reasoning: Semantics are unclear (fuzzy!),but has proved very useful for commercialapplications