Agent Planning

In EURACE, for now, the agents are planning (considering one period in the future) their current activities. This is an individual planning, not a priori.

Stuff to do for May

• Have to work on a possibility for using neural network implementation in place of classifier systems in the bargaining model by Moulet et al.
• Code an example of agents playing Monopoly with rational and dumb agents.
• Read up on Reinforcement Learning and gaming.

Econophysics

Seems game theory is not very promising to be able to do strategic decisions. These startegies are not a big issue as is learning in agents. Econophysics on the other hand seems good.
Coming from physics, an old passion of mine, it focuses on heterogenous agents interacting in non equilibrium situations in economics. Examples include using mechanical equilibrium to explain general equilibrum theory, or using the Black-Scholes formula for pricing is similar to heat diffusion. Wikipedia although does say that the subject is not very favoured by economists.

Mathematical issues

Reading through von Neumann's book (page 47 onwards) games have taken a more mathematocal approach:
If game G has n players. Then for each player there will be a number of moves M1, M2,...,Mv .
For each move there would also exist alternates for moves a(1), a(2),...a(v).
These vary according to the moves or strategies being chosen.

Then comes a second scenario of a game of Poker P. The moves being played will differ is cheating is involved, one player known the moves or cards of the other players, OR if nobody is cheating.

Boring day....

Been reading through some game theory books. Not getting any where yet..