TITLE:
Randomized Constraint Limit Linear Programming in Risk Management
AUTHORS:
Dennis Ridley, Abdullah Khan
KEYWORDS:
Pedagogic Effectiveness of Big Data Analytics, Linear Programming, Stochastic Optimization, Constraint Limit, Profit Distribution and Risk, Monte Carlo Simulation
JOURNAL NAME:
Journal of Applied Mathematics and Physics,
Vol.8 No.11,
November
30,
2020
ABSTRACT: Traditional linear program (LP) models are deterministic. The way that constraint limit uncertainty is handled is to compute the range of feasibility. After the optimal solution is obtained, typically by the simplex method, one considers the effect of varying each constraint limit, one at a time. This yields the range of feasibility within which the solution remains feasible. This sensitivity analysis is useful for helping the analyst get a feel for the problem. However, it is unrealistic because some constraint limits can vary randomly. These are typically constraint limits based on expected inventory. Inventory may fall short if there are overdue deliveries, unplanned machine failure, spoilage, etc. A realistic LP is created for simultaneously randomizing the constraint limits from any probability distribution. The corresponding distribution of objective function values is created. This distribution is examined directly for central tendencies, spread, skewness and extreme values for the purpose of risk analysis. The spreadsheet design presented is ideal for teaching Monte Carlo simulation and risk analysis to graduate students in business analytics with no specialized programming language requirement.