Code to investigate the effects of the Random Forest Hyper-Parameters in the surrogate modeling of multi-objective combinatorial optimization problems
Source code and supplementary material for the manuscript "Effects of the Random Forests Hyper-Parameters in Surrogate Models for Multi-Objective Combinatorial Optimization: A Case Study using MOEA/D-RFTS" published in the IEEE Latin America Transactions.
The datasets for the benchmark problems are in the folder datasets.
The folder supplementary contains the results of the experiments described in the paper.
To perform your own optimization, run the script optimization.py.
The results of the optimization procedures are saved on the folder
tests/results.
To investigate the performance of a Random Forest with a specific set of hyper-parameter values on the datasets, run the script rf_tunning.py.
The results are saved on the folder results/tunning.
Any comments, suggestions or doubts, feel free to contact Matheus Bernardelli de moraes at m121214@dac.unicamp.br