Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Quantum inspired high dimensional hyperparameter optimization of machine learning model
The hyper-parameter optimization of machine learning model is not a completely solved problem. The exquisite combination of artificial tuning and grid search may be a good choice in the area where the dimension of hyper-parameters is very low. But for high-dimensional hyper-parameter optimization problems, artificial tuning and grid search are obviously helpless. In this paper, we propose a quantum inspired hyper-parameter search method, which employs a quantum inspired evolutionary algorithm integrated with the traditional neural network (NN) training technique to evolve a NN subject to accuracy on test set as the objective. We evaluate the proposed method on MNIST dataset and compare it with random search algorithm. Experimental results demonstrate the superiority of our proposed method.
Quantum inspired high dimensional hyperparameter optimization of machine learning model
The hyper-parameter optimization of machine learning model is not a completely solved problem. The exquisite combination of artificial tuning and grid search may be a good choice in the area where the dimension of hyper-parameters is very low. But for high-dimensional hyper-parameter optimization problems, artificial tuning and grid search are obviously helpless. In this paper, we propose a quantum inspired hyper-parameter search method, which employs a quantum inspired evolutionary algorithm integrated with the traditional neural network (NN) training technique to evolve a NN subject to accuracy on test set as the objective. We evaluate the proposed method on MNIST dataset and compare it with random search algorithm. Experimental results demonstrate the superiority of our proposed method.
Quantum inspired high dimensional hyperparameter optimization of machine learning model
Li, Yangyang (Autor:in) / Lu, Gao (Autor:in) / Zhou, Linhao (Autor:in) / Jiao, Licheng (Autor:in)
01.09.2017
380720 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
| DOAJ | 2023