I need help with how I should approach a forecasting problem. I have two tables like the ones below.
Current Amount Not Moved: Table 1 | Past History Items moved by days: Table 2
Days_In Value | Days_In_before_Moved Value Percent
0 1000 | 1 400 2%
1 980 | 2 700 3.5%
4 1100 | 3 350 1.75%
6 500 | 4 600 3%
. . | . . .
. . | . . .
. . | . . .
59 1200 | 60 1400 7%
Estimated Value: Y Total Value: 20000
I'm trying to use the table to the right, which has the avg value of the items that were moved out based on how long they were in the other table, and the percent of the total value that was moved after X amount of days. Currently, I have tried to Add 1 to the days in table 1, then take the percent total from table 2, multiple the percent by value in table 1 to get an expected amount that would be moved, then sum up those values. When I test this on past data, there is always at least a 10% difference between my estimate and the actual amount for that day. Though I was pretty much trying to fit a probability density function from table 2 to table 1, but it doesn't seem to work.
The Days in table 1 are not always consistent, it does have some missing days. I'm thinking about trying to fill those days with some values, but I'm not sure if that is a good approach. Really not sure what approach to take that would be the best. Any ideas would be appreciated.