Comparative Study of Bubble Growth Dynamics with DeepONet

Collaborators: Vivek Oommen, Sayan Chakraborty

As part of the final project requirement for CSCI 2470 (Deep Learning), I worked in a team of three to implement a deep learning model called DeepONet that is able to accurately predict bubble growth dynamics. It is an operator network that is capable of operator approximation in contrast to the more general functional approximation usually performed by deep learning models.

We perform a comparative study between the DeepONet model vs LSTMs, GRUs and Seq2Seq models.

We reference and implement the work by Chensen Lin et al., ‘Operator learning for predicting multiscale bubble growth dynamics’. In this paper, the authors propose a new deep learning framework, the DeepONet, for predicting the multiscale bubble growth dynamics governed by the Rayleigh-Plesset equation.

Complete project documentation along with a demo video and PPT presentation can be accessed here : https://devpost.com/software/comparative-study-of-bubble-growth-dynamics-with-deeponet?ref_content=user-portfolio&ref_feature=in_progress