Wonders of Seabed: Difficulty Evaluation of Management Games Using Neural Network - Technologies and Applications of Artificial Intelligence - page 169

Fig. 3. The evaluation method using neural network. E i ( i =1 , 2 , 3 , 4) are the events
and T j ( j =1 , 2 , 3 , 4) are thresholds. rand returns a probability value.
3. If celebration fee is bigger than T cele , it has chance to occur a happy citizen
event . A happy citizen event leads to that the income for the government is
doubled for this month.
4. If the human resource fee is less than T human , it has a chance to occur a
wrong policy event . A wrong policy event would make a loss of a quarter of
the income.
The percentage of the chance is decided by the value of the parameters (see
Figure 4). One may employ a sophisticated probability model or based on a real
statistics to compute the probability for an event to occur. This is beyond the
scope of this paper.
In the second step, we multiply each input with the weight and add bias.
Then, the result is fed as input to the neural network. After that, we use a
discriminant function to get the outcomes of the current developed city.