Advanced Module : Machine Learning
 • Reference: EMCL-A-ML
 • Courses:
• ML - Machine Learning
• ADNE - Learning with Non-Structured Data
• Requirements:
To obtain the 12 crs of the module, students should obtain 6 crs in each of the two courses.
• Description:
The module aims to provide the students with theoretical and practical knowledge on machine learning, ranging from the basic concepts, the characterisation of data and their normalisation and regularisation, the different paradigms for machine learning, including supervised, unsupervised, reinforcement and deep learning, as well as the techniques that implement them.

-------------------------------------
 ML - Machine Learning
• Reference: 11157
• Semester: Autumn
• Web Page: aa.ssdi.di.fct.unl.pt
• Lecturers: Ludwig Krippahl, Susana Nascimento
• Description:
Both the Supervised and Unsupervised Learning components of the syllabus cover all the core subjects and most of the elective subjects for Machine Learning in the ACM Computer Science Curricula 2013 (http://cs2013.org/).
The syllabus overviews machine learning paradigms (supervised, unsupervised and reinforcement learning), topics of data normalisation and visualisation and several techniques adopted for the different paradigms, including Regression, Artificial Neural Networks, Support Vector Machines and Ensemble methods, (supervised), partitional, probabilistic and fuzzy clustering as well as Markov chains (unsupervised).

-------------------------------------
LNSD - Learning with Non-Structured Data
• Reference: 12084
• Semester: Spring
• Web Page: adne.ssdi.di.fct.unl.pt
• Lecturers: Ludwig Krippahl, Rui Pimenta Rodrigues
• Description:
This course studies basic principles of deep learning. It overviews different data types and problems, and techniques to learn from them, namely deep feedforward networks and their optimisation and regularisation, convolution problems, recurrent networks and sequential problems, an introduction to generator models and their application to unsupervised learning, as well as semi-supervised learning, representation learning and transference learning..