no code implementations • 3 Feb 2023 • Jian Cheng Wong, Pao-Hsiung Chiu, Chinchun Ooi, My Ha Dao, Yew-Soon Ong
On the other hand, if the samples are too sparse, existing PINNs tend to overfit the near boundary region, leading to incorrect solution.
no code implementations • 15 Dec 2022 • Nicholas Sung Wei Yong, Jian Cheng Wong, Pao-Hsiung Chiu, Abhishek Gupta, Chinchun Ooi, Yew-Soon Ong
Hence, neuroevolution algorithms, with their superior global search capacity, may be a better choice for PINNs relative to gradient descent methods.
no code implementations • 29 Oct 2021 • Pao-Hsiung Chiu, Jian Cheng Wong, Chinchun Ooi, My Ha Dao, Yew-Soon Ong
In this study, novel physics-informed neural network (PINN) methods for coupling neighboring support points and their derivative terms which are obtained by automatic differentiation (AD), are proposed to allow efficient training with improved accuracy.
no code implementations • 20 Sep 2021 • Jian Cheng Wong, Chinchun Ooi, Abhishek Gupta, Yew-Soon Ong
In this paper, we present a novel perspective of the merits of learning in sinusoidal spaces with PINNs.
no code implementations • 5 May 2021 • Jian Cheng Wong, Chinchun Ooi, Pao-Hsiung Chiu, My Ha Dao
In addition, we propose a novel transfer optimization scheme for use in such surrogate modeling scenarios and demonstrate an approximately 3x improvement in speed to convergence and an order of magnitude improvement in predictive performance over conventional Xavier initialization for training of new scenarios.