Sensitivity analysis
We conducted two types of sensitivity analysis.
1. Screen Upper Bound
In order to figure out how much the result will be affected when the upper bound for the number of screens is modified, we ran solver table on constraint #3, which is the upper bound for the number of screens. The input cell was the maximum number of screens, and the output cells were total slack and movie schedule. We varied the maximum number from 1 to 8. According to the result from the solver table, the optimal solution turned out to be 4, showing the least slack. Currently set upper bound, 3, involved second least total slack. This meant that current distribution was not that bad considering that total slack showed only 2 in difference. As for the distribution of movies to each screen, the current one was very similar to that of optimal solution. In screen 3 and 4, they differed just in order, and in screen 6, current distribution had one less movie.

2. Total Expected Demand
We assumed certain amount of demand, but that assumption can vary according to unexpected situation. Therefore we conducted the second sensitivity analysis on total expected demand. Setting current demand as 100%, we varied the amount from 50%, which is half the current amount, to 300%, which is 3 times the current amount, with the increment of 10%. In solver table, the input cell was the total expected demand and the output cells were movie distribution and total slack. We ran the solver table and the optimal solution turned out to be 120% of current demand. On the table, as the total demand changed, the number of screens for each movie varied sensitively. The current demand incurred fourth least total slack. Of note, 110% case showed the same number of movies and 130% case showed one less movie than 100% case, while maintaining less total slack. This can be interpreted as the efficiency