2. OUTLINE Course description, What is pattern recognition, Cost of error, Decision boundaries, The desgin cycle

3. The objective of the course is to make student familiar with general approaches such as – Bayes classification, – discriminant functions, – decision trees, – nearest neighbor rule, – neural networks for pattern recognition. Course description

4. Course Book: – R.O. Duda, P.E. Hart, and D.G. Stork, Pattern Classification, New York: John Wiley, 2001, Course description

5. WEEKLY SCHEDULE AND PRE-STUDY PAGES Wee k TopicsPre-study Pages 1IntroductionChapter 1 (main text) 2Bayesian Decision TheoryChapter 2 3Bayesian Decision TheoryChapter 2 4Bayesian Decision TheoryChapter 2 5Maximum – Likelihood and Bayesian Parameter EstimationChapter 3 6Maximum – Likelihood and Bayesian Parameter EstimationChapter 3 7Nonparametric TechniquesChapter 4 8Nonparametric TechniquesChapter 4 9Linear Discriminant FunctionsChapter 5 10Linear Discriminant FunctionsChapter 5 11Multilayer Neural NetworksChapter 6 12Nonmetric MethodsChapter 8 13Unsupervised Learning and ClusteringChapter 10 14Unsupervised Learning and ClusteringChapter 10

6. EVALUATION SYSTEM IN-TERM STUDIESQUANTITYPERCENTAGE Mid-terms22x20=40 Assignment330 Final Exam130 TOTAL100 Course description

7. What is a pattern ? an entity, vaguely defined, that could be given a name, e.g.: – fingerprint image, – handwritten word, – human face, – speech signal, – DNA sequence,

8. What is pattern recognition ? a discipline which learn some theories and methods to design machines that can recognize patterns in noisy data or complex environment (Srihari,Govindaraju).

9. What is pattern recognition ? a scientific discipline whose aim is the classification of the objects into a lot of categories or classes. Pattern recognition is also a integral part in most machine intelligence system built for decision making (Sergios Theodoridis).

10. What is pattern recognition ? the act of taking in raw data and making an action based on the category of the pattern (Duda and Hart)

11. What is pattern recognition ? Pattern recognition is the study of how machines can: – observe the environment, – learn to distinguish patterns of interest, – make sound and reasonable decisions about the categories of the patterns.

12. What is pattern recognition ? Some Applications:

13. What is pattern recognition ? Some Applications:

14. What is pattern recognition ? Some Applications:

15. What is pattern recognition ? Some Applications:

16. What is pattern recognition ? Some Applications:

17. What is pattern recognition ? Some Applications:

18. An Example Suppose that: – A fish packing plant wants to automate the process of sorting incoming fish on a conveyor belt according to species, – There are two species: Sea bass, Salmon.

19. An Example How to distinguish one specie from the other ? (length, width, weight, number and shape of fins, tail shape,etc.)

20. An Example Suppose somebody at the fish plant say us that: – Sea bass is generally longer than a salmon Then our models for the fish: – Sea bass have some typical length, and this is greater than that for salmon.

21. An Example Then length becomes a feature, We might attempt to classify the fish by seeing whether or not the length of a fish exceeds some critical value (threshold value) l*.

22. An Example How to desie on the critical value (threshold value) ?

23. An Example How to desie on the critical value (threshold value) ? – We could obtain some training samples of different types of fish, – make length measurements, – Inspect the results.

24. An Example Measurement results on the training sample related to two species.

25. An Example Can we reliably seperate sea bass from salmon by using length as a feature ? Remember our model: – Sea bass have some typical length, and this is greater than that for salmon.

26. An Example From histogram we can see that single criteria is quite poor.

27. An Example It is obvious that length is not a good feature. What we can do to seperate sea bass from salmon?

28. An Example What we can do to seperate sea bass from salmon? Try another feature: – average lightness of the fish scales.

29. An Example Can we reliably seperate sea bass from salmon by using lightness as a feature ?

30. An Example Lighness is better than length as a feature but again there are some problems.

31. An Example Suppose we also know that: – Sea bass are typically wider than salmon. We can use more than one feature for our decision: – Lightness (x 1 ) and width (x 2 )

32. An Example Each fish is now a point in two dimension. – Lightness (x 1 ) and width (x 2 )

33. An Example Each fish is now a point in two dimension. – Lightness (x 1 ) and width (x 2 )

34. An Example Each fish is now a point in two dimension. – Lightness (x 1 ) and width (x 2 )

35. Cost of error Cost of different errors must be considered when making decisions, We try to make a decision rule so as to minimize such a cost, This is the central task of decision theory.

36. Cost of error For example, if the fish packing company knows that: – Customers who buy salmon will object if they see sea bass in their cans. – Customers who buy sea bass will not be unhappy if they occasionally see some expensive salmon in their cans.

37. Decision boundaries We can perform better if we use more complex decision boundaries.

38. Decision boundaries There is a trade of between complexity of the decision rules and their performances to unknown samples. Generalization: The ability of the classifier to produce correct results on novel patterns. Simplify the decision boundary!

39. The design cycle

40. Collect data: – Collect train and test data Choose features: – Domain dependence and prior information, – Computational cost and feasibility, – Discriminative features, – Invariant features with respect to translation, rotation and scale, – Robust features with respect to occlusion, distortion, deformation, and variations in environment.

41. The design cycle Choose model: – Types of models: templates, decision-theoretic or statistical, syntactic or structural, neural, and hybrid. Train classifier: – learn the rule from data, Evaluate classifier: – estimate the performance – problems of overfitting and generalization.

42. References Srihari, S.N., Covindaraju, Pattern recognition, Chapman &Hall, London, 1034-1041, 1993, Sergios Theodoridis, Konstantinos Koutroumbas, pattern recognition, Pattern Recognition,Elsevier(USA)),1982 R.O. Duda, P.E. Hart, and D.G. Stork, Pattern Classification, New York: John Wiley, 2001,