How knowledge is distributed
in the population?
Mayby!!
Contents
First and foremost, the distribution of information and knowledge should have something to do with learning. So I start with an exponential learning curve, which is easy to implement.
Simple learning curve
It does not seem realistic that the learn rate will remain constant over time. Therefore, the learning rate is redefined on the basis of current knowledge. In the absence of better knowledge, a factor of 0.5 is used.
with updated learn rate
The basic idea is that when two agents meet, they learn together. Later, this should happen in a network. In the beginning, I will let the agents meet randomly in the population to see if the implementation of joint learning works.
Simulating random Meetings
If it works as it should, it will be expanded so that a certain percentage of the population meets at the same time.
Grouped in Slots
A daily structure with a certain number of working hours is introduced to enable better interpretation of the results.
… in a Day Structure
Now it seems to be time to add other aspects. At first it seems central that different areas of knowledge should be possible.
Areas of Knowledge
Up to now, the topic has been defined in terms of knowledge; now it should be a question of preference.
… with prefernces
The time has come for the next big step! So far, the encounters have been purely random, but this is about to change. A first assumption is that the agents are pure utility maximisers of their preferred knowledge.
Simulation of selected Meetings
It seems utopian to believe that each agent knows exactly how much knowledge all other agents have, even if they work at different universities. For this reason, an ingroup and an outgroup are defined for the utility calculation, so that outside the university only the average knowledge of the university is known.