Individual
and collective learning in multi-agent Systems
1 - Markets as
cognitive and interactive complex systems
2.3. Collective learning (3) : Statistical
physics approach of markets (a)
Statistical physics provide atoolbox
of concepts & methodssuitables for to
analyze the agregate regularity of systems of interacting
elements, given a microscopic description of its elements
Physicists (especialy in the field of
condensed matter) are increasingly interested by to apply
their tools outside their field, sometime in mixed teams
: evolution of biological species(Bak-Sneppen,
) , immunology (Solomon - Atlan), social
and political decision making
(Galam-Moscovici) market organization
(Bak Chen, Scheinkman, Woodford 1993; Kirman, Weisbuch,
1995, 1996, 1998; Bak, Paczuski, Shubick 1997 ; Stauffer
& alii 2000).
A large group of Physicists are
developping by their own a special interest for financial
markets : stock, exchange, options and futures this
new field of activity is call by this scholars :
« econophysics » (Bouchaud, Potters, 1997 ;
Kertesz, Kondor (eds.), 1998; Mantega, Stanley, 2000.
« Econophysics » is defined
by Cont (1999) as a new field at the crossroad of
statistical physics, economics and probability theory.