@inproceedings{scholars10686, publisher = {Institution of Engineering and Technology}, year = {2018}, journal = {IET Conference Publications}, title = {Optimization of neural network hyperparameters for gas turbine modelling using bayesian optimization}, volume = {2018}, note = {cited By 7; Conference of 5th IET International Conference on Clean Energy and Technology, CEAT 2018 ; Conference Date: 5 September 2018 Through 6 September 2018; Conference Code:144672}, number = {CP749}, url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85061389406&partnerID=40&md5=2d4c01dae42b2ba4ac2e638f80f47927}, abstract = {Gas turbine model can be used for many applications that can improve gas turbine operation{\^a}??s efficiency and reliability. Artificial Neural Network (ANN) was identified as a good technique for gas turbine modelling. However, ANN is sensitive to its hyperparameters. Current approaches for optimizing the hyperparameters of ANN used in gas turbine modelling has high computational cost. Hence, this paper proposes optimizing the ANN hyperparameters using Bayesian optimization. Bayesian optimization was used to determine the near-optimum number of layers and number of neurons for the developed gas turbine model. The result shows hyperparameters optimization using Bayesian optimization results neural network with better prediction accuracy compared to hyperparameters optimization using random search. {\^A}{\copyright} 2018 Institution of Engineering and Technology. All rights reserved.}, isbn = {9781785618161; 9781785618437; 9781785618468; 9781785618871; 9781785619427; 9781785619694; 9781839530036; 9781785617911}, author = {Tarik, M. H. M. and Omar, M. and Abdullah, M. F. and Ibrahim, R.}, keywords = {Gases; Neural networks; Optimization, Bayesian; Bayesian optimization; Computational costs; Efficiency and reliability; Gas turbine modeling; Hyper-parameter; Prediction accuracy; Turbine operation, Gas turbines} }