A Regularized Physics-Informed Neural Network to Support Data-Driven Nonlinear Constrained Optimization

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

1. In the abstract, please highlight the novel contributions and advances within the research area.

2. The training strategy is required to presented in detail, such as the initial learning rate... 

3. In section 6.2, there are three evaluation scenarios are provided to validate the performance of regularisation functions in data generation. The results report that the proposed work is better because it is more stable and less affected by outliers in the third case. In fact, a notable limitation of our approach is that its effectiveness may decrease in scenarios where the data lacks specific details at the attribute level.

4. For a better analysis, please also provide the time and space complexity of the proposed neural network.

 

Author Response

See attached pdf.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The authors present framework for supporting data-driven nonlinear constraint optimization. The authors report that the use of the principles of physics leads to competitive performance compared to existing solutions. Experimental results are provided for two scenarios involving supervised and unsupervised datasets.

 Strengths:

-        A strong literature review.

-        The paper is well-written, and the methods are clearly presented.

Minor comments:

-        The choice of the parameters in Subsection 5.1 is not explained. Have you evaluated any other architectures?

-        The authors are suggested to add an error analysis section for better understanding the limitations of the model.

Author Response

See attached pdf.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

All problems have been addressed.