Università di Pavia - Offerta formativa

DEEP LEARNING

Anno immatricolazione

2020/2021

Anno offerta

2021/2022

Normativa

DM270

SSD

ING-INF/05 (SISTEMI DI ELABORAZIONE DELLE INFORMAZIONI)

Dipartimento

DIPARTIMENTO DI INGEGNERIA INDUSTRIALE E DELL'INFORMAZIONE

Corso di studio

COMPUTER ENGINEERING

Curriculum

Data Science

Anno di corso

2°

Periodo didattico

Secondo Semestre (07/03/2022 - 17/06/2022)

Crediti

6

Ore

45 ore di attività frontale

Lingua insegnamento

English

Tipo esame

ORALE

Docente

PIASTRA MARCO (titolare) - 6 CFU

Prerequisiti

Foundations of linear algebra and multivariable calculus. Practical experience with at least one programming language. Some acquaintance with Python and Numpy.

Obiettivi formativi

The course follows a conceptual pathway that starting from simple linear regression to the sophisticated aspects of state-of-art of deep convolutional neural networks, deep recurrent networks and deep reinforcement learning. A unifying mathematical approach is followed throughout this path, to encompass and make it possible to understand the basic features of modern software frameworks for deep learning, such as TensorFlow.

Programma e contenuti

1) Deep Supervised Learning

Algebraic model, foundations of tensor calculus

Learning as representation, evaluation and optimization

Single-layer networks as universal approximators

Dataset in tensor representation for calculus

Flow diagrams, automatic differentiation

Regression and classification, softmax

Deep layered representation, modularity

Ottimizzazione

2) Deep Convolutional Neural Networks

Convolutional layers and complex architectures

Data augmentation and Transfer learning

Layered learning, different optimization processses

Fallibility and adversarial models

Classification, object detection, segmentation

3) Deep Recurrent Networks

Temporal unfolding, shared-parameters layers

Long-Short Term Memory (LSTM)

4) Deep Reinforcement Learning

On-policy and off-policy learning

Actor critic and advance function

Neural MCTS: AlphaZero e MuZero

Algebraic model, foundations of tensor calculus

Learning as representation, evaluation and optimization

Single-layer networks as universal approximators

Dataset in tensor representation for calculus

Flow diagrams, automatic differentiation

Regression and classification, softmax

Deep layered representation, modularity

Ottimizzazione

2) Deep Convolutional Neural Networks

Convolutional layers and complex architectures

Data augmentation and Transfer learning

Layered learning, different optimization processses

Fallibility and adversarial models

Classification, object detection, segmentation

3) Deep Recurrent Networks

Temporal unfolding, shared-parameters layers

Long-Short Term Memory (LSTM)

4) Deep Reinforcement Learning

On-policy and off-policy learning

Actor critic and advance function

Neural MCTS: AlphaZero e MuZero

Metodi didattici

Lectures (hours/year in lecture theatre): 30

Practical class (hours/year in lecture theatre): 16

Practicals / Workshops (hours/year in lecture theatre): 0

Practical class (hours/year in lecture theatre): 16

Practicals / Workshops (hours/year in lecture theatre): 0

Testi di riferimento

See the home page of the course (http://vision.unipv.it/DL) for lecture slides, suggested readings and software for the exercises.

Modalità verifica apprendimento

The evaluation includes the realization of a project, to be agreed in advance. The final exam is an interview about the theory, together with the discussion of the project.

Altre informazioni

Obiettivi Agenda 2030 per lo sviluppo sostenibile