@inproceedings {pub4253,
	title = {Transfer Learning for Gaussian Process Assisted Evolutionary Bi-objective Optimization for Objectives with Different Evaluation Times},
	author = {Xilu Wang AND Sebastian Schmitt AND Markus Olhofer AND Yaochu Jin},
	year = {2020},
	abstract = {Despite the success of evolutionary algorithms (EAs) for
solving multi-objective problems, most of them are based on
the assumption that all objectives can be evaluated within
the same period of time. However, in many real-world ap-
plications, such an assumption is unrealistic since diff erent
objectives must be evaluated using diff erent computer simu-
lations or physical experiments with various time complex-
ity. To address this issue, a surrogate assisted evolutionary
algorithm along with a parameter-based transfer learning
(T-SAEA) is proposed in this work. While the surrogate for
the cheap objective is updated by suffi cient training data
due to the more available evaluations, the surrogate for the
expensive one is updated by either the training data set or
a transfer learning approach. To fi nd out the transferable
knowledge, a fi lter-based feature selection is used to capture
the pivotal features of each objective, making the common
features as the dependencies between the diff erent objectives.
Then, the corresponding parameters in surrogate models are
adaptively shared to facilitate the effi ciency of the surrogate
modeling. The performance of the proposed algorithm is com-
pared with fi ve state-of-the-art algorithms on the DTLZ and
UF benchmark suites respectively. The experimental results
demonstrate that the proposed algorithm outperforms the
compared algorithms on the multi-objective optimization
problems for objectives with diff erent evaluation times.},
	publisher = {ACM},
	booktitle = {GECCO 2020}
}