Teaching code

4S009018

Credits

6

Coordinator

Marco Cristani

Language

English

Also offered in courses:

• Machine Learning & Deep Learning of the course Master's degree in Artificial intelligence
• Machine Learning & Deep Learning of the course Master's degree in Artificial intelligence

Scientific Disciplinary Sector (SSD)

INF/01 - INFORMATICS

Moodle Seminars 0

The teaching is organized as follows:

Learning objectives

The course provides the theoretical foundations and describes advanced methodologies related to the area of deep learning. Deep learning solves machine learning and pattern recognition problems through the neural network paradigm and the numerical optimization. The course is also highly implementative, offering specific programming concepts for professional Python-based systems. Deep learning aims to build mainly nonlinear regression and classification systems based on neural networks. A neural network can be seen as a simple computational structure (multinomial logistic functions + non-linearities, intertwined together) which is enhanced by structuring itself at various levels (layers) of various types (fully connected, convolutional, recurring, and many others). Each of these structures underlies a very precise theory (for example the dropout of neural networks refers to the Bayesian approximation) which will be formally detailed by the teacher. In this way, the student will not only be a user of the discipline, but will manage it by acquiring formal critical skills. Particular attention will also be paid to the aspect of explainability, that is, all those techniques capable of communicating critical cases in which a particular neural network is unable to solve problems. The course will provide case studies on which to apply the studied techniques, to make them immediately usable in a professional context.

Prerequisites and basic notions

Probability and Statistics
Image Processing

Program

The course presents a series of state-of-the-art topics in the field of recognition. Deep learning will be analyzed starting from its methodological basis. Each topic will be explained through updated articles together with the lesson slides. The following books are suggested as a reference:
- Christopher M. Bishop. 2006. Pattern Recognition and Machine Learning (Information Science and Statistics). Springer-Verlag New York, Inc., Secaucus, NJ, USA.
- Ian Goodfellow, Yoshua Bengio, and Aaron Courville. Deep learning. MIT Press, 2016.
Topics:
- Linear Regression, ridge, LASSO, elastic net
- Multinomial Logistic Classifier,
- Neural Networks,
- Backpropagation,
- Convolutional Neural Network,
- Recurrent Neural Networks
- Long Short-Term Memory machine
- Transformer Network

Bibliography

Didactic methods

The lessons will be in the presence, any supplementary material can be recovered from the video lessons of the previous year

Learning assessment procedures

The exam involves the discussion with the teacher of a written paper, which proposes a solution to an industrial classification problem. The grade will depend on the classification skills obtained by the classifier (with different measures of classification goodness from problem to problem), on the statistical confidence margins offered and on the theoretical motivation that prompted the student to choose a particular programming technique.

Evaluation criteria

Ability to face a recognition problem with the most appropriate tools. Ability to motivate implementation choices with theoretical notions. Ability to discuss results with interpretability

Criteria for the composition of the final grade

15/30 quality of the project. 15/30 capacity for appropriate discussion of the project itself

Exam language

Inglese