Abstract
To adapt to the environment it necessary that intelligent machines must have the capability to learn. This chapter presents the basic concepts and techniques of learning found in humans, as well as their implementation aspects for machines. The chapter presents the challenges of building learning capabilities in machines, types of machine learning, and the relative efforts needed to build these learning capabilities. The philosophy of the discipline of machine learning is presented. The basic model of learning is discussed, followed by the classes of learning—supervised and unsupervised—then various techniques of inductive learning—argument based learning, online concept learning, propositional and relational learning, and learning through decision trees—are presented in sufficient details. Other techniques like discovery-based learning, reinforced learning, learning and reasoning through analogy, explanation-based learning are presented, with some worked examples. Finally, the potential applications of machine learning, the basic research problems in machine learning, followed by chapter summary, and a set of exercises are appended.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
Instance and example are the same here, which are input to the function.
References
Ahmadi M et al (2007) IFSA: incremental feature-set augmentation for reinforcement learning tasks. In: The sixth international joint conference on autonomous agents and multi-agent systems, pp 1128–1135
Bengio Y (2016) Machines who learn. Sci Am 6:38–43
Domingos P (2012) A few useful things to know about machine learning. Commun ACM 55(10):78–87
Mantaras RL, Armengol E (1998) Machine learning from examples: inductive and Lazy methods. Data Knowl Eng 25(1998):99–123
Mozina M et al (2007) Argument based machine learning. Artif Intell 171:922–937
Schmid U, Kitzelmann E (2011) Inductive rule learning on the knowledge level. Cogn Syst Res 12:237–248
Sunstein CR (1993) On analogical reasoning. Harv Law Rev 106:741–791
Tadepalli P, OK D (1998) Model-based average reward reinforcement learning. Artif Intell 100:177–224
Wang J, Gasser L (2002) Mutual online concept learning for multiple agents. In: AAMAS’02, July 15–19, Bologna, Italy
Winston PH (1980) Learning and reasoning by analogy. Commun ACM 23(12):689–703
Zupan B et al (1999) Learning by discovering concept hierarchies. Artif Intell 109:211–242
Author information
Authors and Affiliations
Corresponding author
Exercises
Exercises
-
1.
At the end of the day, try to recollect any five important events you have witnessed this day. Try to associate the type of learning you have used in these events.
-
2.
Analyze and propose the type of structures you consider are appropriate in the following learning processes:
-
a.
Rote learning
-
b.
Inductive learning
-
c.
Supervised learning
-
d.
Unsupervised learning
-
e.
Learning by example
-
f.
Analogical learning
Answer the above questions, in reference to some programming language, e.g., C. And, describe the estimated algorithmic steps. For example, for rote learning you may take the example of the number tables, and indexing you need, so that one can speak the table of say “5”, and also one can answer \(5 \times 6 = 30\).
-
a.
-
3.
What is the difference between the learning by induction versus learning by examples.
-
4.
Explain the major difference between analogical learning and inductive learning in respect of the approach used for learning.
-
5.
Suggest a learning method for each of the following, explaining why the suggested method is appropriate, and provide logical steps to learn using it.
-
a.
To learn how to drive a car after having observed a trained driver, while you are riding along with the same.
-
b.
To learn how to drive a car after having known the driving of a bullock-cart.
-
c.
To learn how to drive a car after having learned the driving of a tractor.
-
d.
To learn how to fly an aircraft after having very closely observed the birds flying, like eagle, crane, and hawk.
-
e.
Learning to keep your wallet protected after having lost it due to theft.
-
a.
-
6.
The helicopter takes off straight, instead of having a run before taking off. It has similarity with peacock in the birds. Explain a mathematical model, which supports the learning by analogy of a copter with a peacock.
Rights and permissions
Copyright information
© 2020 Springer Nature India Private Limited
About this chapter
Cite this chapter
Chowdhary, K.R. (2020). Machine Learning. In: Fundamentals of Artificial Intelligence. Springer, New Delhi. https://doi.org/10.1007/978-81-322-3972-7_13
Download citation
DOI: https://doi.org/10.1007/978-81-322-3972-7_13
Published:
Publisher Name: Springer, New Delhi
Print ISBN: 978-81-322-3970-3
Online ISBN: 978-81-322-3972-7
eBook Packages: Computer ScienceComputer Science (R0)