154
X-Machines for Agent-Based Modeling: FLAME Perspectives
do nothing-promotions-hire worker-hire worker-make redundant-sell
goods-assess production
Here the mall tries to hire more workers in its strategy, we see a more
sharper fall in capitals. Although the mall switches to selling as the Mall
1 did, it cannot recover most of its losses, and goes bankrupt sharply.
Interestingly, also self-evolving Mall 3 started with Mall 2’s behavior,
but did not adopt it for long. This is probably due to people being
employed by Mall 2 and no more applications were left. Also there are
not many sales for Mall 2. Mall 2 runs before in the agent order during
the simulation. Thus Mall 3 tries out new strategies but keeps closing
down. Because of so many inactive periods, the mall in unable to climb
up in its sales and soon goes bankrupt.
Best performing people groups. There were three groups of people which
have the same information within the group. It was up to the individuals
to use information. The gene length of people functions was five and
included:
1. Find job.
2. Buy cheap goods.
3. Buy advertised.
4. Buy recommended.
5. Buy random.
6. Do nothing.
7. Quit job.
Figure 6.19(b) shows that Group 1 and 2 are able to compete with each
other, whereas Group 3 can not. Savings are largely influenced by wages
of the people. Figure 6.20(a) shows the different wages of the people.
As the wages of people in Group 2 are higher, they are able to save up
more money than Group 1.
Since Mall 2 goes bankrupt at an early stage, it does not show any
impact on people savings. But when Mall 3 goes bankrupt at t ≈ 125,
we see a considerable fall in the gradient of savings. The demand wage
decreases as more and more people become jobless and try to get new
jobs with other existing malls.
Learning windows of the people. In Figures 6.19(b) and 6.20(a), Group
2 is doing better than Group 1. This is due to their learning windows
being smaller, so people in Group 2 are quicker to react and change
behavior and adapt better.
X-Machines for Agent-Based Modeling: FLAME Perspectives
do nothing-promotions-hire worker-hire worker-make redundant-sell
goods-assess production
Here the mall tries to hire more workers in its strategy, we see a more
sharper fall in capitals. Although the mall switches to selling as the Mall
1 did, it cannot recover most of its losses, and goes bankrupt sharply.
Interestingly, also self-evolving Mall 3 started with Mall 2’s behavior,
but did not adopt it for long. This is probably due to people being
employed by Mall 2 and no more applications were left. Also there are
not many sales for Mall 2. Mall 2 runs before in the agent order during
the simulation. Thus Mall 3 tries out new strategies but keeps closing
down. Because of so many inactive periods, the mall in unable to climb
up in its sales and soon goes bankrupt.
Best performing people groups. There were three groups of people which
have the same information within the group. It was up to the individuals
to use information. The gene length of people functions was five and
included:
1. Find job.
2. Buy cheap goods.
3. Buy advertised.
4. Buy recommended.
5. Buy random.
6. Do nothing.
7. Quit job.
Figure 6.19(b) shows that Group 1 and 2 are able to compete with each
other, whereas Group 3 can not. Savings are largely influenced by wages
of the people. Figure 6.20(a) shows the different wages of the people.
As the wages of people in Group 2 are higher, they are able to save up
more money than Group 1.
Since Mall 2 goes bankrupt at an early stage, it does not show any
impact on people savings. But when Mall 3 goes bankrupt at t ≈ 125,
we see a considerable fall in the gradient of savings. The demand wage
decreases as more and more people become jobless and try to get new
jobs with other existing malls.
Learning windows of the people. In Figures 6.19(b) and 6.20(a), Group
2 is doing better than Group 1. This is due to their learning windows
being smaller, so people in Group 2 are quicker to react and change
behavior and adapt better.