
Course unit
COMPUTATIONAL NEUROSCIENCE
SCP8082718, A.A. 2019/20
Information concerning the students who enrolled in A.Y. 2018/19
ECTS: details
Type 
ScientificDisciplinary Sector 
Credits allocated 
Educational activities in elective or integrative disciplines 
INGINF/05 
Data Processing Systems 
6.0 
Course unit organization
Period 
First semester 
Year 
2nd Year 
Teaching method 
frontal 
Type of hours 
Credits 
Teaching hours 
Hours of Individual study 
Shifts 
Lecture 
6.0 
48 
102.0 
No turn 
Examination board
Board 
From 
To 
Members of the board 
1 A.A. 2019/2020 
01/10/2019 
15/03/2021 
TESTOLIN
ALBERTO
(Presidente)
CHIARIOTTI
FEDERICO
(Membro Effettivo)
ERSEGHE
TOMASO
(Supplente)
ZANELLA
ANDREA
(Supplente)
ZORZI
MICHELE
(Supplente)

Prerequisites:

The course relies on preliminary knowledge of mathematical analysis, linear algebra and probability theory. Familiarity with machine learning concepts is desired, though not mandatory. Python programming skills are required. 
Target skills and knowledge:

The course covers the theory and practice of artificial neural networks, highlighting their relevance both for artificial intelligence applications and for modeling human cognition and brain function. Theoretical discussion of various types of neural networks and learning algorithms is complemented by handson practices in the computer lab (PyTorch framework). 
Examination methods:

Evaluation of knowledge and abilities acquired will consist on an individual project assignment, which will be discussed during the oral exam. The project will require a software implementation of one or more computational models and analyses discussed during the course, along with a short essay in which the student will describe and discuss the project results. The oral exam will also include general theoretical questions related to the course content. 
Assessment criteria:

The evaluation is based on the understanding of course topics and the acquisition of the proposed concepts and methodologies. 
Course unit contents:

1. Introduction: computational and mathematical modeling of neural systems; basics of neuroscience; levels of analysis in system neuroscience.
2. Singleneuron modeling: morphology, neuroelectronics, principles of synaptic transmission; integrateandfire models; the HodgkinHuxley model.
3. Principles of neural encoding: recording neuronal responses; spike trains, firing rates, local field potentials; tuning functions and receptive fields; efficient encoding principles and information compression.
4. Network modeling: neural network architectures; localistic, distributed, and sparse representations; examples from the visual system.
5. Learning, memory and plasticity: synaptic plasticity in biological systems (Hebb rule, LTP, LTD, STDP); synaptic plasticity in artificial neural networks and overview of machine learning basics.
6. Supervised learning: perceptron, delta rule, error backpropagation.
7. Supervised deep learning: advanced optimization methods for training multilayer networks; convolutional architectures; transfer learning and multitask learning.
8. Recurrent neural networks: backpropagation through time, long shortterm memory networks.
9. Unsupervised learning: competitive networks; selforganizing maps; associative memories and Hopfield networks; autoencoders and Boltzmann machines.
10. Unsupervised deep learning: hierarchical generative models; generative adversarial networks.
11. Reinforcement learning: explorationexploitation dilemma; temporaldifference learning; conditioning and dopamine circuits; deep reinforcement learning.
12. Case studies from neurocognitive modeling: visual perception; space coding; semantic cognition; complementary learning systems; hippocampus and experience replay.
13. Largescale brain organization: structural and functional properties of brain networks; neuronal oscillations and spontaneous brain activity; neuromorphic hardware. 
Planned learning activities and teaching methods:

Teaching is based on frontal lectures covering the theory, and practice classes on neural network modeling in the computer lab. Interactive teaching techniques will be used, including thinkpairshare and interactive discussions of a few minutes on open questions. This will enforce interactive learning and the ability to critically reflect on the concepts discussed. 
Additional notes about suggested reading:

All topics will be covered during the lectures. Slides will be made available on elearning. Students' notes must be integrated with the reference books and with further material (mostly scientific articles) provided by the teacher on the elearning platform. 
Textbooks (and optional supplementary readings) 

Goodfellow, I., Bengio, Y., and Courville, A., Deep Learning. : MIT Press, 2016. Electronic version freely available online

Dayan, P., and L. F. Abbott, Theoretical neuroscience. : MIT Press, 2001. Electronic version freely available online

Hertz, J., Krogh, A., and Palmer, R. G., Introduction To The Theory Of Neural Computation. : Westview Press, 1991.


