Syllabus

COURSE STRUCTURE 

The course will have a problem-based structure for each lesson as follows:

- Lectures: where theoretical aspects will be discussed (generally 1h per day).

- Recitations/Tutorials: Computer projects are investigated collectively. You will learn from practical challenges (generally 1h per day). We will decrease every week the hours of python. 

OUTCOMES

At the end of this course, you will be able to

1. Master and explain fundamental statistical concepts related to prediction of ‘events’.

2. Write programs to perform statistical prediction or pattern recognition.

3. Be familiar with related topic useful to Pattern recognition: Optimization and dimensionality reduction. You will be capable of seeing how aspects of these topics influence pattern recognition.

4. Understand how the chosen features a models influence the outcomes of the algorithms.

5. Relate and apply pattern recognition algorithms to everyday problems (e.g. image analysis and bioinformatics).

ATTENTIVE COURSE OUTLINE WITH DATES

16/03 Introduction  (1h)

- DHS Ch 1, Bishop Ch. 1.2.

- Practical/tutorial work (1h)

17/03 Statistics review and Introduction to optimization and Image (1h)

 - Practical/tutorial work (1h)

18/03 Bayesian decision theory. (1h)

- DHS Ch 2, Bishop Ch. 1.5.

- Practical/tutorial work (1h)

19/03 The curse of Dimensionality Reduction, outline Sparsity. (1h)

- Bishop Ch. 12.1, 4.1.4.

- On the Role of Sparse and Redundant Representations in Image Processing, M. Elad et al., IEEE on Applications of Sparse Representation & Compressive Sensing, 2010.

- practical/tutorial work (1h)

20/03 Artificial Neural Networks (Perceptrons, backpropagation).  (1h)

- Bishop Ch. 5.1-5.4.

- practical/tutorial work (1h)

--------------------------------------------------------------------------------------

23/03 Supervised and Unsupervised Learning, Self-Organizing Maps.

-Self-Organized Formation of Topologically Correct Feature Maps. T. Kohonen. Biological Cybern. 1982.

- Optimization, more details.

24/03 Maximum likelihood estimation. (1h)

- Bishop Ch. 2.3.

- practical/tutorial work (1h)

25/03 K-Means and K-NN.  (1h)

- Bishop 9.1

- practical/tutorial work (1h)

26/03 Discriminant functions and Support vector machine.

- Bishop Ch. 7.1.

- practical/tutorial work (1h)

27/03 Generalization and Measuring test errors.   (1 h)

- HTF Ch. 7-8.

- An introduction to ROC analysis, T. Fawcett, Pattern Recognition Letters  2006.- practical/tutorial work (1h )

- My slides on p-values

- practical tutorial

---------------------------------------------------------------------------------------------------------------------

30/03 Non-linear dimensionality reduction and Kernel methods pt 1. (1h) 

- A Global Geometric Framework for Nonlinear Dimensionality Reduction, J. Tenenbaum et al.,Science 2000.

- Nonlinear Dimensionality Reduction by Locally Linear Embedding, S. Roweis and L. Saul, Science  2000.

- practical/tutorial work (1h)

31/03:   Businespplan review 1h

- practical tutorial 1h

Image processing and machine learning on Ultrasound images 1h

01/04 Businessplan day

Co-working 1h

Local Linear embedding 1h

02/03  Non-linear dimensionality reduction and Kernel methods pt 2. (1h) 

- Bishop Ch. 6.

- practical/tutorial work (1h)

03/04 OpenCV (image and shape analysis in c++/Android). 

Only lab activities, no report.1h

Presentations day