After completing this course, the student:
- knows important methods of statistical pattern recognition and their theoretical foundation
- knows general principles of statistical learning, such as over-fitting and the bias-variance decomposition
- knows how to apply methods of statistical pattern recognition to practical data analysis problems
- can perform an experimental evaluation of statistical learning methods in a sound manner
- knows the fundamental deep learning topics
- has practical skills for applying deep learning methods (e.g. feed forward neural networks, CNNs, RNNs)
- can model a research-oriented problem in the application domain of deep learning and suggest a solution using deep learning techniques
The course is graded through a practical assignment (20% of the final mark), a group project (40%) and a written exam (40%).
A repair test requires at least a 4 for the original test.
In this course we study statistical pattern recognition and machine learning.
The subjects covered are:
- General principles of data analysis: overfitting, bias-variance trade-off, model selection, regularization, the curse of dimensionality.
- Linear statistical models for regression and classification.
- Clustering and unsupervised learning.
- Support vector machines.
- Neural networks and deep learning.
Knowledge of elementary probability theory, statistics, multivariable calculus and linear algebra is presupposed.
Lectures and computer lab sessions.
- Book: Christopher M. Bishop, Pattern Recognition and Machine Learning, Springer, 2006.
- Book: Ian Goodfellow, Yoshua Bengio and Aaron Courville, Deep Learning, MIT Press, 2016.
Book URL: http://www.deeplearningbook.org.
- Possibly additional literature in the form of research papers, book chapters, etcetera.