Machine Learning

Recommended Prerequisites

BSc in Informatics Engineering or equivalent.

Teaching Methods

Theoretical classes , 2h weekly, with audiovisual and computational means. Computational demonstrations studied techniques for machine learning.

Practical classes, for the development of mini-projects covering the several themes of the syllabus. Each mini-project occupies in average 2,5 classes and is developed in groups of 2 or 3 students. Some of the mini-projects have a research component, the students are challenged to search on the recent literature ideas for their implementation. The Matlab+Simulink + Toolboxes, and/or Weka are used for computer implementations.

Learning Outcomes

To study the main techniques of machine learning in the context of the multiplicity of data types available in practical applications. Namely, studies include the techniques such as decision trees, artificial neural networks, fuzzy logic, fuzzy and neuro-fuzzy systems, configuration and training of support vector machines. To develop competencies to design systems for classification of large data sets, for diagnosis in industrial and medical contexts, for intelligent control, for holistic analysis of complex problems and critical evaluation of its results. Additionally, competencies for group working, for scientific and technical oral and written communication are developed. Generic competencies in analysis and synthesis, informatics knowledge relative to the study focus, problem solving, critical thinking, decision capability, autonomous learning, practical application of theoretical knowledge, creativity, self-criticism and self-evaluation, and research.     

Work Placement(s)

No

Syllabus

Chap.1 Introduction to machine learning

Chap.2 Decision trees

Chap.3 Artificial Neural Networks: basic architectures

Chap.4. Multilayer NN and training by retropropagation

Chap.5. RBF Neural networks.Training algorithms.

Chap.6. Advanced NN architectures. Recursive NN and their learning algorithms

Chap.7. Fuzzy logic, fuzzy sets, fuzzy relations and Zadeh extension principle.

Chap.8. Fuzzy rule based systems of Mamdani and Sugeno types. Learning of fuzzy rules and training fuzzy systems: clustering techniques.

Chap.9.Neuro-fuzzy systems: the ANFIS architecture and its training. Applications.

Chap. 10. Suppport Vector Machines.

Head Lecturer(s)

António Dourado Pereira Correia

Assessment Methods

Assessment
Resolution Problems: 20.0%
Project: 30.0%
Exam: 50.0%

Bibliography

Foundations of Machine Learning , Mehryar Mohri, Afshin Rostamizadeh and Ameet Talwalkar MIT Press, 2012

Pattern Recognition and Machine Learning, C.M. Bishop,Springer 2006

Neural Network Design, Hagan, Demuth and Beale, PWS Publishing, 1995.

Neural Network Toolbox Users´s Guide, The Mathworks, 2012

Fundamentals of Artificial Neural Networks, Hassoun. M. H.,MIT Press, 1994.

Neural and Adaptive Systems, J.C. Príncipe, N.R. Euliano, W. C. Lefevre, Wiley, 2000

Fuzzy Logic With Engineering Applications, 2nd Ed., Timothy Ross, McGraw Hill, 2004.

Fuzzy Logic Toolbox Users´s Guide, The Mathworks, 2012.

Introduction to Neuro-Fuzzy Systems, Robert Fullér, Springer Verlag 2000.

Neural Networks: A Comprehensive Foundation,Simon Haykin,Prentice Hall,1999

Fuzzy Modelling and Control, Andrzej Piegat, Springer Verlag, 2001.