An efficient contesting procedure for AutoML optimization

Nguyen, D.A.; Kononova, A.V.; Menzel, S.; Sendhoff, B.; Bäck T.H.W. (2022)

Article / Letter to editor

Automated Machine Learning (AutoML) frameworks are designed to select the optimal combination of operators and hyperparameters. Classical AutoML-based Bayesian Optimization (BO) approaches often integrate all operator search spaces into a single search space. However, a disadvantage of this history-based strategy is that it can be less robust when initialized randomly than optimizing each operator algorithm combination independently. To overcome this issue, a novel contesting procedure algorithm, Divide And Conquer Optimization (DACOpt), is proposed to make AutoML more robust. DACOpt partitions the AutoML search space into a reasonable number of sub-spaces based on algorithm similarity and budget constraints. Furthermore, throughout the optimization process, DACOpt allocates resources to each sub-space to ensure that (1) all areas of the search space are covered and (2) more resources are assigned to the most promising sub-space. Two extensive sets of experiments on 117 benchmark datasets demonstrate that DACOpt achieves significantly better results in 36% of AutoML benchmark datasets: 5% when to compared to TPOT, 8% - to AutoSklearn, 15% - to H20 and 18% - to ATM.

• All authors: Nguyen, D.A.; Kononova, A.V.; Menzel, S.; Sendhoff, B.; Bäck T.H.W.
• Date: 2022-07-18
• Journal: IEEE Access
• Volume: 10
• Pages: 2169 - 3536
• DOI: doi:10.1109/ACCESS.2022.3192036
• Link: https://ieeexplore.ieee.org/document/9832605
• Sponsorship: Horizon 2020(H2020)
• Grant number: 766186