Machine Learning and? Neural Networks by? Pascual Campoy pascual.campoy@upm.es Computer Vision Group Universidad Politécnica Madrid 1 table of contents What is it about? Objectives Topics Scheduling Evaluation Methodology Learning material Class-room exercises 2
What is it about? intelligence machine intelligence machine learning artificial neural networks dimensionality reduction data mining pattern recognition 3 Objectives: Be able to explain following terms and relate them to real-word problems: - learning, pattern recognition, classification, dimensionality reduction, supervised learning, unsupervised learning Ability to apply classical PR techniques - PCA, Bayesian Decision & K-means Ability to apply supervised ANN: - MLP to PR Ability to apply unsupervised ANN: - SOM to PR 4
Topics: 1. Intelligence & learning 2. Feature processing 3. Classical classifiers 4. Machine Learning 5. Bio-inspiration 1st. Part (exam and work) 6. Supervised ANN: Multilayer Perceptron 7. Non supervised ANN: Self-organized Maps 2nd. Part (exam and work) 5 Schedule Schedule 2010-2011 for "Pattern Recognition & Neural Networks" Topic Subject Week/day 0 Guide to the subject 15-S 1 Intelligence & Learning 22-S 2 Feature processing 29-S, 6-O 3 Classical Classifiers 13-O 4 Machine learning 20-O 5 Bio-inspiration 27-O 6 a Supervised Neural Networks: MLP 3-N 1-2-3-4 Presentation Parctical Work #1 10-N 1-2-3-4 1st Exam 17-N 6 b Supervised Neural Networks: MLP 24-N 7 Unsupervised Neural Networks: SOM 1-D 4-5-6-7 Presentation Parctical Work #2 15-D 2nd. Exam 31-J 6
schedule 3 ECTS x 25 hours/ects = 75 hours - Classroom 28 hours: 14 weeks x 2 hours/week - Outside the classroom 47 hours: 47 hours/14 weeks = 3,5 hours/week 7 Evaluation Continuous evaluation - Class-room exercises - Two practical works - Two exams 1 4 5 Momentary evaluation - Practical works (compulsory) - Oral exam 1,5 8,5 8
Methodology Practice: Roger Schank interviewed by Eduardo Punset Collaborative: SCALE-UP http://scaleup.ncsu.edu/ classroom at MIT 9 methodology In the collaborative classroom: - Lecture - Collaborative working on the computer - Tutorial - Presentation of practical works Out of the classroom: - Individual study (bofore-after) - Two Practical Works In the informatics classroom: - Two individual exams 10
Learning material Aulaweb: invited student pr53000038, password learning - This guide: 0_MLaNN_guide.pdf - Slides for every topic, including classroom exercices - Two practical works - Dataset for exercises and pracical works, including exercise form exercises_template.doc 11 learning material Books: "Pattern Classification" Duda-R, Hart-P, Stork-D Wiley-Interscience, 2004 Pattern recognition & Machine Learning Christopher M. Bishop Springer, 2006 12
Further reading biological inspiration Christof Koch Rodolfo Llinás V.S. Ramachandran 13 further reading making things to work Jeff Hawkins David Fogel 14
Class-room exercices 1. Create a word document from the template exercises_template.doc (downloaded from Aulaweb/contenidos/problemas) with the name: ex_yy_aaaaa_bbbbb_ccccc.doc where X is the chapter number YY is the exercise number within the chapter AAAAA, BBBBB and CCCCC are the students ID numbers 2. Save this document in your PC at Documentos compartidos/entregar 15 class-room exercices 3. Fulfill the heading: 4. Write in the document the required solution that includes the explanation, the Matlab code, the obtained graphics and results, and the comments and conclusions 5. The document has to be closed in order to allow to be collected 16
Example 0.1 change Matlab working directory to Documentos_ compartidos/entregar download datos_d2_c3_s1 from Aulaweb into this directory >> load datos_d2_c3_s1 p.valor 2x1000 p.clase 1x1000 p.salida 1x1000 >> plot(p.valor(1,:),p.valor(2,:), b. ); hold on; 17 Exercise 0.1 >> load datos_d2_c3_s1.mat Print p.valor using a diferent color/prompt for each clase determined in p.clase 18
What is PR by learning?? 20