关键词: 深度卷积神经网络, 分类模型, 超参数优化, 代理模型, 模型优化

Abstract: To further improve the efficiency of hyperparameter multi-objective adaptive optimization of deep classification models, a Filter Enhanced Dropout Agent model （FEDA）was proposed. Firstly, a dual-channel Dropout neural network with enhanced point-to-point mutual information constraint was constructed, to enhance the fitting of high-dimensional hyperparameter deep classification model, and the selection of candidate solution sets was accelerated by combining the aggregation solution selection strategy. Second, an algorithm FEDA-ARMOEA combined with model management strategies was designed to balance the convergence and diversity of individual populations, and to assist FEDA in improving the efficiency of deep classification model training and hyperparameter self optimization. Comparative experiments were conducted between FEDA-ARMOEA, EDN-ARMOEA, HeE-MOREA, and other algorithms. Experimental results show that FEDA-ARMOEA performs well on 41 sets of testing problems. Experiments on industrial application weld data set MTF and public data set CIFAR-10 show that the accuracy of FEDA-ARMOEA optimized classification model on MTF and CIFAR-10 data set is 96.16% and 93.79%, respectively, and the training time is decreased by an average of 34.29% and 25.55% compared with the contrast algorithms, respectively. All of them are superior to the contrast algorithms', which verifies the effectiveness and generalization of the proposed algorithm.

Key words: deep convolutional neural network, classification model, hyperparameter optimization, agent model, model optimization