Additive regularization schedule for neural architecture search
This addresses neural architecture search efficiency for machine learning practitioners, though it appears incremental as it builds on existing regularization concepts.
The paper tackles neural network structure optimization by proposing an additive regularization schedule for neural architecture search, where regularizers corresponding to different network parts are scheduled during optimization. Computational experiments show the method finds efficient, low-complexity networks with improved accuracy compared to non-regularized models.
Neural network structures have a critical impact on the accuracy and stability of forecasting. Neural architecture search procedures help design an optimal neural network according to some loss function, which represents a set of quality criteria. This paper investigates the problem of neural network structure optimization. It proposes a way to construct a loss function, which contains a set of additive elements. Each element is called the regularizer. It corresponds to some part of the neural network structure and represents a criterion to optimize. The optimization procedure changes the structure in iterations. To optimize various parts of the structure, the procedure changes the set of regularizers according to some schedule. The authors propose a way to construct the additive regularization schedule. By comparing regularized models with non-regularized ones for a collection of datasets the computational experiments show that the proposed method finds efficient neural network structure and delivers accurate networks of low complexity.