2. Outline
• Learning
• Machine Learning
• Inductive Learning
• Why Inductive Learning
• Inductive Learning Methods
• Rules Family Of Algorithms
• Application of inductive learning
3. Learning
An agent is learning if it improves its performance on
future tasks after making observations about the world.[1]
A computer program is said to learn from experience E
with respect to some class of tasks T and performance
measure P, if its performance at tasks in T, as measured
by P, improves with experience E .[2]
4. Why would agent learn?
1. The designers cannot anticipate all possible
situations that the agent might find itself in
2. The designers cannot anticipate all changes over
time
3. Sometimes human programmers have no idea how to
program a solution themselves
5. Machine Learning
Machine learning is concerned with computer
programs that automatically improve their performance
through experience.[1]
Field of study that gives computers the ability to learn
without being explicitly programmed. [3]
7. Inductive Learning
• A new field of machine learning known as inductive
learning has been introduced to help in inducing general
rules and predicting future activities.[4]
• Inductive learning is learning from observation and earlier
knowledge by generalization of rules and conclusions.
Inductive learning allows for the identification of training
data or earlier knowledge patterns.[5]
• The identified and extracted generalized rules come to use
in reasoning and problem solving.[6]
9. Inductive learning
• Inductive:
This cat is black. That cat is black A third cat is black. Therefore
all cats are black
• Deductive:
Bachelor's are unmarried men. Bill is unmarried. Therefore, Bill
is a bachelor
10. Why inductive learning
Alternative method of knowledge acquisition in which
knowledge learned or induced from examples
• Human experts are capable of using their knowledge in their
daily work, but they usually cannot summaries and
generalize their knowledge explicitly in a form which is
sufficiently systematic, correct and complete for machine
representation and application .[7]
• While it is very difficult for an expert to articulate his
knowledge, it is relatively easy to document case studies of
the expert's skills at work.[7]
11. Inductive Learning Methods
1-Divide-and- conquer
• CLS
• ID3
• C4.5
2-Covering
• AQ family
• CN2
• RULES
A set of training examples usually used to form a decision
tree.[5]
12. Rules Family Of Algorithms
Rule1
• Pham and Aksoy have developed RULES-1 (RULe Extraction
System-1).[8]
• The first member of RULES family of algorithms RULES-1.[9]
• Extracts rules for objects in similar sets of classes
• Each object has its own attributes and values
• The attributes and the values associated with them in a collection of
objects form an array of attributes and values
• Total number of elements of the array is the total number of all
possible values
• EX: 4 attributes with values (3,4,2,5).
Total number of elements is 14
13. Rules Family Of Algorithms
Rule1
Example:
• Attributes: Weather, Temperature
• Values:{rainy, sunny, snowy} and {low, average, high}
• Attribute-value pairs:
(weather,rainy),(weather,sunny),(weather,snowy),(temperature,low),(
temperature,average),(temperature,High)
15. Rules Family Of Algorithms
Rules versions:
Rule-1
Rule-2
Rule-3
Rule-3 plus
Rule-4
Rule-5
Each version has some extra new features to overcome some problems
that cannot be coped with using previous versions. The algorithms have
been used for many application which shows their good performance .[8]
16. Application of inductive learning
• The technology for building knowledge-based systems by
inductive inference from examples has been demonstrated
successfully in several practical applications .[10]
• Inductive learning algorithms are domain independent and
can be used in any task involving classification or pattern
recognition.[11]
Making Credit Decisions
Education
Medical applications
17. References
• [1] S.R. RUSSELL and P.Norvig , “Artificial Intelligence: A Modern Approach”, 3d.ed. USA,
Prentice Hall,2009, pp. 693-767.
• [2] slide2[2] T. M. Mitchell ,“Machine Learning”,1st.ed. U.K, McGraw-Hill ,1997,pp. 1-5.
• [3] A.L.Samuel ,”Some Studies in Machine Learning Using the Game of Checkers”, vol.11
, no.6,pp. 601 - 617,Nov. 1967.
• [4] H. A. ELGIBREEN and M. S. AKSOY,“RULES – TL : A SIMPLE AND IMPROVED
RULES”,J. Theor. Appl. Inf. Technol., vol. 47, no. 1, 2013.
• [5] A. M. AlMana and M. S. AKSOY, “An Overview of Inductive Learning Algorithms” ,” Int.
J. Comput. Appl., vol. 88 – no.4, 2014.
• [6] A. H. Mohamed and M. H. S. Bin Jahabar, “Implementation and Comparison of Inductive
Learning Algorithms on Timetabling”, Int. J. Inf. Technol., vol. 12, no. 7, pp. 97–113, 2006.
• [7] M.S.Aksoy “A review of inductive learning algorithms”, Journal of Faculty of Management,
Istanbul, Vol.25, No.2, pp.171-186, Turkey, 1996.
• [8] M. S. Aksoy, “A Review of RULES Family of Algorithms”,Math. Comput. Appl., vol. 13, no.
1, pp. 51–60, 2008.
• [9] D.T.Pham and M.S.Aksoy, “RULES: A simple rule extraction system”, Expert Systems with
Applications, Vol.8, No.1, pp.59-65, USA, 1995.
• [10] J.R. Quinlan, ”Induction of Decision Trees”,vol 1, Centre for Advanced Computing
Sciences, New South Wales Institute of Technology, Sydney , pp 81-106 ,2007.
• [11] M. S. Aksoy, A. Almudimigh, O. Torkul, and I. H. Cedimoglu, “Applications of Inductive
Learning to Automated Visual Inspection”, Int. J. Comput. Appl., vol. 60, no. 14, pp. 14–18,
2012.