Cogmaster, Course CA6a

(course homepage here)

This page will be regularly updated.

Rem.: some documents can be accessed only with the login and password given to registered students;

these documents are provided as course support, and cannot be distributed or used for anything else without authorization.
**Class 6**, Nov. 16: Learning I.
**Unsupervised learning**

**Class 7**, Nov. 23: Learning II. Behavioral learning

(course homepage here)

This page will be regularly updated.

Rem.: some documents can be accessed only with the login and password given to registered students;

these documents are provided as course support, and cannot be distributed or used for anything else without authorization.

This page:

General and useful references: see bottom of this page.

Specific material for the Course, current Academic year: see below

*Introduction and basic tools, models and concepts*

**Class 1**, Oct 5:

**Class 12**, Oct 6: (Biophysics of) Neurons II.

**Class 3**, Oct 19: Synapses.

**Class 4**, Oct 26: Balanced networks.

**Class 5**, Nov. 11: Rate models

Introduction

(Biophysics of) Neurons I: Basic electrical properties and simplest model

presentation: slides (pdf)

lecture notes (taken by students, academic year 2015-2016).

Hodgkin & Huxley: access to papers here.

presentation: slides (pdf)

lecture notes (taken by students, academic year 2015-2016).

related papers here.

Rate models: slides (pdf)

Papers: dynamics

(*Wilson and Cowan 1972; Seung 2003; Murphy and Miller 2009*)

Introduction

Different kinds of learning: supervised, unsupervised and reward learning.

Supervised learning: learning associations

Hebbian learning with binary synapses: the Willshaw model.

Supervised learning by the Perceptron: capacity, information storage.

Perceptron: Papers

Hertz, Krogh, and Palmer (1991) (see general references below), Chap. 5.

The Oja model

presentation: slides (pdf)

E. Oja, A simplified neuron model as a principal components analyzer, J. Math. Biol. 15 (1982) 267-273.

Hertz, Krogh, and Palmer (1991) (see general references below), Chap. 8.2 pp. 199-210

lecture notes (taken by students, academic year 2015-2016).

Papers here. See in particular:

- Kaelbling et al, 1996

*(an exhaustive survey on Reinforcement Learning and Dynamic Programming methods for Markov Decision Processes)*

- Sutton, 1988

*("the" TD learning paper)*

- Grandman et al, 2012

*(a somewhat technical survey on actor-critic methods)*

- Bromberg-martin et al, 2010

*(a recent survey of dopamin signaling and its role in learning)*

- Schultz, Dayan and Mantague, 1997

*(the pioneering study showing that DA neuron resemble a reward prediction error)*

- Gallistel, 2003

*(a stimulating paper on classical conditioning)*

- Sutton, 1988

- Grandman et al, 2012

- Bromberg-martin et al, 2010

- Schultz, Dayan and Mantague, 1997

- Gallistel, 2003

Presentation: slides (pdf)

More course material on neural coding here.

Presentation: slides (pdf)

Presentation: slides (pdf)

Presentation: slides (pdf)

- academic year 2016/2017: see here.
- academic year 2015/2016: see here.
- academic year 2014/2015: see here.

Back to top of page (Course material) - Back to Course homepage.

GENERAL REFERENCES

*Most, if not all,
the following books are available at the **RISC library
and at the library of Institut Henri Poincaré.*

*
Computational neuroscience books

Abbott
and Dayan, "Theoretical Neuroscience" (MIT Press, 2001)

A nice overview of the field of computational neuroscience, at a level comparable to that of the course.

Gerstner, Kistler, Naud and Paninski,
"Neuronal Dynamics: From single neurons to networks and models of cognition" (Cambridge University Press, July 2014)

Narrower scope, but a nice online version.

Ermentrout and Terman,
"Mathematical Foundations of Neuroscience" (Springer, 2010)

A book with detailed mathematical treatments, with a focus on the dynamics of neural activity.

Tuckwell,
"Introduction to Theoretical Neurobiology", Vols. I & II (Cambridge
U. Press, 1988)

Again mathematical treatment, with narrower scope and emphasis on single-neuron dynamics and statistics of neural activity.

Koch,
"Biophysics of Computation" (Oxford U. Press, 1999)

*A
detailed presentation of computational aspects of single-neuron behavior,
closer to biology than Tuckwell's book (but more restricted in topics also).*

Hertz,
Krogh, and Palmer, "Introduction to the Theory of Neural Computation" (Addison-Wesley,
1991 - now from: Perseus Book Group and Westview Press)

*A
good reference book on memory models in the perspective of statistical
physics (and on many other neural network models). This book is more advanced
mathematically than the three other and a good resource for learning useful
statistical physics techniques.*

Rieke,
Warland, de Ruyter van Steveninck, and Bialek, "Spikes" (MIT Press, 1997)

*A
nice introduction to the statistical and information theoretic approach
to neural coding, with a title length inversely proportional to the number
of authors.*

Izhikevich, "Dynamical Systems in Neuroscience: The Geometry of Excitability and
Bursting" (MIT Press, 2007)

*
General neurobiology books:

Nicholls,
Martin, Wallace, and Fuchs, "From Neurons to Brain" (Sinauer Associates,
2001, 4th ed.)

*A
readable and well-balanced introductory survey of neurobiology.*

Kandel,
Schwartz, and Jessel, "Principles of Neural Science" (McGraw-Hill, 4th
edition, 2000)

*Another
successful neurobiology survey text.*

Hubel,
"Eye, Brain, and Vision" (Scientific American Library, 1988 and 1995)

*A
delightful introduction to visual processing by the brain, at an undergraduate/interested
general reader level.*

Johnston
and Wu, "Foundations of Cellular Neurophysiology" (MIT Press, 1994)

*A
comprehensive introduction to the biophysics of neurons, with a quantitative
leaning.*

Fain,
"Molecular and Cellular Physiology of Neurons" (Harvard University Press, 1999)

*A nice very read on the physiology of single neurons and synapses, at an introductory level, with a lovely final part on
sensory transduction.*

*
Periodicals:

A number of journals are devoted to computational neuroscience; these include Neural Computation, The Journal of Computational Neuroscience, Network: Computation in Neural Systems, Biological Cybernetics, Neural Networks. Other general or topical science journals also publish computational neuroscience articles; prominently, Nature, Nature Neuroscience, Science, Neuron, PNAS, PloS (free, online), The Journal of Physiology, The Journal of Neuroscience, The Journal of Neurophysiology, Cerebral Cortex, Vision Research, Trends in Neurosciences, Current Opinion in Neurobiology.

Back to top of page (Course material) - Back to Course homepage: French/ English version.