Improving Multi-fidelity Optimization with a Recurring Learning Rate for Hyperparameter Tuning
This work addresses hyperparameter tuning inefficiencies for deep learning practitioners, offering an incremental improvement over existing multi-fidelity methods.
The paper tackles the problem of efficiently exploring hyperparameter search spaces for CNNs by addressing the slow convergence of high-performing configurations in multi-fidelity optimization, proposing MORL which improves low-fidelity approximation and outperforms methods like SHA and Hyperband while achieving significant gains over hand-tuned configurations within practical budgets.
Despite the evolution of Convolutional Neural Networks (CNNs), their performance is surprisingly dependent on the choice of hyperparameters. However, it remains challenging to efficiently explore large hyperparameter search space due to the long training times of modern CNNs. Multi-fidelity optimization enables the exploration of more hyperparameter configurations given budget by early termination of unpromising configurations. However, it often results in selecting a sub-optimal configuration as training with the high-performing configuration typically converges slowly in an early phase. In this paper, we propose Multi-fidelity Optimization with a Recurring Learning rate (MORL) which incorporates CNNs' optimization process into multi-fidelity optimization. MORL alleviates the problem of slow-starter and achieves a more precise low-fidelity approximation. Our comprehensive experiments on general image classification, transfer learning, and semi-supervised learning demonstrate the effectiveness of MORL over other multi-fidelity optimization methods such as Successive Halving Algorithm (SHA) and Hyperband. Furthermore, it achieves significant performance improvements over hand-tuned hyperparameter configuration within a practical budget.