Studieretningsprojekt - Neurale Netværk

Programme af dagen

Fredag d. 23. november 2012 (uge 47)


10:15	Workshoppen starter
12:30	Frokost i kantinen
15:30	Workshoppen ender

Litteratur

Slides from the meeting

Kernestof

[B1] S. Russell and P. Norvig, 2010. Artificial Intelligence: A Modern Approach. Third Edition. Prentice Hall, sec. 18.7, pp. 727-737 (photocopies)

Andre kilder

[B2] J. Glenn Brookshear, 2009. Computer Science - An Overview, 10th edition, Pearson, sec. 11.5 pp. 550-560. (photocopies)

[B3] Kandel E, Schwartz J, Jessel T (2000) Principles of Neural Science. 4th edition. New York: McGraw-Hill Medical.
(Good reference for the biological account of the brain.)

[B4] David Kriesel, 2007, A Brief Introduction to Neural Networks,

[B5] A.M. Turing. Computing Machinery and Intelligence. Mind, Oxford University Press on behalf of the Mind Association, 1950, 59(236), 433-460. http://dx.doi.org/10.1093/mind/LIX.236.433

[B6] G.F. Luger. Artificial Intelligence: Structures and Strategies for Complex Problem Solving. Addison-Wesley, 2009

[B7] Simon Haykin. Neural networks: A comprehensive foundation. 842 pages. 2 edition. Prentice Hall. 1998
(In depth treatment from a machine learning perspective)

[B8] Christopher M. Bishop Neural Networks for Pattern Recognition, Oxford Univ. Press. (In depth treatment from a machine learning perspective, it assumes statistical knowledge)

[B9] D.E. Rumelhart, B. Widrow and M.A. Lehr. The basic ideas in neural networks. Commun. ACM, ACM, 1994, 37(3), 87-92 (short article reviewing the way artificial networks work. Accessible)

[B10] Gazzaniga, M.S., Ivry, R., & Mangun, GR (2002) Cognitive Neuroscience: The Biology of the Mind. 3rd Edition. New York: W.W. Norton.

[L1] Rodney Brook's Lecture

Software

Neural Networks Demos at AIspace
Demo on Gradient Method
Perceptron
LeNet 5 Neural network for handwriting recognition

Kontekst

Neuroscience (follow the link to Scholarpedia entry)

is concerned with how the biological nervous systems of humans and other animals are organized and how they function
clinically oriented neuroscience aimed at better understanding and treating neurological and other diseases by means of improved molecular pharmacology.
Computational neuroscience is a subfield of neuroscience that uses mathematical methods to simulate and understand the function of the nervous system.

Cognitive Psychology

Like traditional neuroscience, psychology is a field that took firm root in the late 19th C. In contrast to the explicitly reductionist goals of most neuroscience, the goal of psychology has been to understand 'mental' functions and behavior as such rather than their anatomical, physiological and molecular bases. Yet, differently from philosophy, psychology is based on experimental evidence.
cognitive psychology has, been devoted to understanding those aspects of human mentation that entail conscious awareness and all that goes with it, as opposed to unconscious processes and behaviors generally described as 'reflexes'.

Cognitive Neuroscience

Cognitive neuroscience joins cognitive psychology with neuroscience since higher order brain functions and behaviors ultimately depend on the same molecular, cellular and circuit level machinery

Artificial Neural Nets

Currently, the term Artificial Neural Network (ANN) tends to refer mostly to neural network models employed in statistics and artificial intelligence.

ANN have their main applications in machine learning, data mining and data forecasting. They are usually used to model complex relationships between inputs and outputs or to find patterns in data.

The approach inspired by biology has been largely abandoned for a more practical approach based on statistics and signal processing.

Examples of appliccations are:


-	recognize numbers in a handwritten ZIP code;
-	estimate the amount of glucose in the blood of a diabetic person from the infrared absorbtion spectrum of that person;
-	identify the risk factors for cancer, based on clinical and demographic variables;
-	classify flowers according to the length of leaves and petals;
-	predict the price of a stock, or the default of a company, in 6 months from now, on the basis of company perfomance measures and economic data;
-	predict the height of water in a river on the basis of historical data.

Artificial Neural Networks are also used as models for simulation/emulation of the central nervous systems in theoretical neuroscience and computational neuroscience.

Forberedelse

I kan læse den Scholarpedia entry og de to uddrager fra B1 og B2, som inholder den matematiske kernstof af workshoppen.

Neural netværk finder anvendelese mellem andre i forudsigelser. Hvis jeres projekt handler om dette, kan i forberede og medbringe de data i vil bruge.

I maa gerne prøve den software fra Software afsnitt.

Deltager


Elev	Gymnasium	Lærer	Interesser
Jehad Barakji	Rungsted Gymnasium	Knud Johnsen	bioteknologi, matematik, cellekommunikation i nervesystemet
Rikke Andersen	Tønder Gymnasium	Søren Widt	matematik, biologi
Cilia Dalgaard	Odense Katedralskole	Per Pedersen	matematik, bioteknologi
Jakob Miland-Samuelsen	Aurehøj gymnasium	Johanne Lund Christiansen	matematik,
Linnea B. Pedersen	Rysensteen Gymnsaium	Søren Schultz Søgaard	matematik
Grete Petersen Fink	Tønder Gymnasium	Jonas Cilieborg

comments powered by Disqus

Date: 2012-11-23 18:00:43 CET

Author: Marco Chiarandini

Org version 7.8.02 with Emacs version 23

Validate XHTML 1.0