A Genetic Algorithm For Fair Land Allocation
Resumo
The goal ofagrarian reform projects is the redistribution offarmland from large latifundia to smaller, often family farmers. One ofthe main problems the Brazilian National Institute ofColonization and Agrarian Reform (INCRA) has to solve is to subdivide a large parcel ofland into smaller lots that are balan- ced with respect to certain attributes. This problem is difficult since it considers several constraints originating from legislation as well as ethical considerati- ons. Current solutions are computer-assisted, but manual, time-consuming and error-prone, leading to rectangular lots ofsimilar areas which are uneven with respect to soil aptitude and access to hydric resources. In this thesis, we propose a genetic algorithm to produce fair land subdivisions automatically. We pre- sent a greedy randomized constructive heuristic based on location-allocation to generate initial solutions, as well as mutation and recombination operators that consider characteristics of the problem. Experiments on real-world and artificial instances confirm the effectiveness of the different components of our method, and show that it leads to more fair solutions than those currently ap- plied in practice.
Referências
Gliesch, A., Ritt, M., and Moreira, M. C. O. (2017). A genetic algorithm for fair land allocation. In Genetic and Evolutionary Computation Conference - GECCO ’17, pages 793–800, New York, New York, USA. ACM Press.
Gliesch, A., Ritt, M., and Moreira, M. C. O. (2018). A Multistart Alternating Tabu Search for Commercial Districting. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), volume 10782 LNCS, pages 158–173.
Hess, S. W., Weaver, J. B., Siegfeldt, H. J., Whelan, J. N., and Zitlau, P. A. (1965). Nonpartisan Political Redistricting by Computer. Operations Research, 13(6):998– 1006.
