Quantitative Forecasting

Part 2: Optimizing solutions

R Batzinger

Optimizing solution

Requires

  • Target: the statistic that summaries the results. This value will be optimized to be the max. min, or equal to a value
  • Variables: this is a sequence of variables originally set to 0’s. The system will spin the numbers to find the setting that is closest to the target.
  • Contraints: This is a list of requirements that must be met.

Simplex Tableau

Uses Excel Solver

  • Uses sumproduct() to calculate capability (capacity * staffnum)

  • The product is compared with the constraint

  • The system works until an optimum is reached.

Adding new shifts

This added feature allows for better tracing of the need giving a lower cost alternative

![simplex2.png]

The Bread Box: a real world example

  • 3 yrs of a Cafe POS register data
  • Includes different types of drink, snacks
  • Useful for management forecasts for staff, sales items inventory, changes caused by holiday
  • A cafe