Optimization of resource allocation and routing in political campaign problems

Abstract

In this thesis, we introduce the problem of a politician who has to allocate his limited budget among meeting and advertising campaigns in the most efficient way so that he can get the maximum number of votes. This problem, optimization of resource allocation and routing on political campaign problems (ORARPCM), assumes that campaigns consist of mass meetings and advertisements that include many types like social media ads, TV ads, party flags or license plate frames. While the politician determines the most effective budget allocation for his campaign, he also selects the cities to run each campaign type and creates the best yielding route for mass meetings according to some predefined distance criteria. In this study, the aim is to observe only the effects of different budget allocations on total vote gain. The proposed ORARPCM can be classified as a multi period, dynamic and time-restricted prize collecting traveling salesman problem (PC-TSP) with budget limitation. Under these assumptions, mathematical model is proposed and numerical results of a Turkey case study is provided. As the solution method, a deterministic dynamic programming model is constructed by using C++. Results show that under the above-mentioned criteria, deterministic dynamic programming is a very efficient solution method for ORARPCM, as the problem is decomposed based on time periods. Besides the computational time is very reasonable. Keywords: Campaign Routing problem, Resource allocation, Dynamic programming, Vehicle routing problem