@article{ozaki2022multiobjective,
title={Multiobjective Tree-Structured Parzen Estimator},
author={Ozaki, Yoshihiko and Tanigaki, Yuki and Watanabe, Shuhei and Nomura, Masahiro and Onishi, Masaki},
journal={Journal of Artificial Intelligence Research},
volume={73},
pages={1209--1250},
year={2022}
}
This repository is no longer maintained.
MOTPE is available on Optuna.
Please use the Optuna's MOTPE implementation (TPESampler).
Please install the following Python packages.
Run MOTPE with the default settings on WFG4.
python main.py
Several commandline options are available.
python main.py --benchmark_name WFG4 --num_objectives 2 --num_variables 9 --k 1 --num_max_evals 250 --num_initial_samples 98 --init_method lhs --num_candidates 24 --gamma 0.10 --seed 128