Product Line Optimization Algorithms

The paper is a part of collaborative research project of Higher School of Economics and «Leroy Merlin». It is dedicated to algoritms of product line optimization. The task is useful in marketing needs and its variations are applied in different business cases. The aim is to develop and compare exsisted algorithms and provide a recomendaion of what algorithm to use. The thing is that the most widely used method is enumeration of combinations, but it is too time consuming and need to be updated. There are four aproaches, which are explored in the paper: enumeration, the greedy algorithm, linear programming and Shapley value. Methods were tested both on random generated data and real data, provided by «Leroy Merlin». The make choice of one the best algorithm wasn't made, because the results are complex and different method is superior to the others in special cases. However, in case of real data experiments the optimal solution was found for any product line size. The conclusion includes some recomendations of pros and cons of each method.