First time here? Checkout the FAQ!
x
+1 vote
597 views
asked in Machine Learning by (140 points)  
What kind of algorithm would best for following problem.
I try to forecast reservation of different kind of tables. Let's say I have 100 different tables, which are reserved for from 17.00-22.00 daily. Each table is either reserved (1) or available (0) on certain hour.  I'm interested in to forecast each table based on history data from 2 previous weeks. So for example result is that tomorrow 18.00-19.00 certain table is either 0 (available) or 1 (reserved).
  

1 Answer

+1 vote
answered by (116k points)  

Before suggesting the best method, you need to explore more on your data in EDA stage. Are these tables dependent to each other or not? Could it be modeled with time-series or not because your problem could be a time-series problem.

The problem you described is similar to hotel room demand forecasting to me and you can review the related papers. You can also find some source codes here for predicting a time-series similar to your problem here.

commented by (140 points)  
Thank you. If I reframe this problem to forecast only a single table. So the result is either 0 or 1. Could it then be a time-series problem?
commented by (116k points)  
It depends on the nature of the problem. What kind of table is that? Are the cells dependant on each other or not? What do you mean by "table"? Is this a Matrix of 0, and 1?
commented by (140 points)  
Thanks for reply. :)
I meant a restaurant table. You can also reframe this how to forecast a single hotel room reservations.
...