Conditional Domain Adaptation with α-Rényi Entropy Regularization and Noise-Aware Label Weighting

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The Research paper provides respectable findings and is well-written, but before it is accepted, it needs to be strengthened in the following ways:

 

The author should compare the CREDA framework to existing methods, and incorporating α-Rényi entropy offers a theoretical advantage.

 

The author should evaluate the performance, stability, and effectiveness of the proposed method, including its use of kernel-based conditional alignment and matrix-based Rényi entropy.

 

The author should evaluate if CREDA's experiments are conducted on standard datasets, if its performance is fair against advanced methods, and if evaluation metrics are clear and appropriate.

 

The author should evaluate the dimensionality reduction and class activation maps, their ability to enhance interpretability, reveal insights into sample relevance, and support semantic consistency.

 

The author should evaluate the accuracy and robustness and provide a quantitative analysis of Rényi entropy regularization.

Comments on the Quality of English Language

The author should ensure that typos and grammar errors are corrected in various places.

Author Response

See attached pdf.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

please see the attachment.

Comments for author File: Comments.pdf

Comments on the Quality of English Language

Suggest Major Revision.

Author Response

See attached pdf.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

This work presents a novel and elegant theoretical framework that combines Rényi entropy with domain adaptation, achieving consistent performance improvements across three standard benchmarks. The proposal stands out for its interpretability component through UMAP and Grad-CAM++ analysis, offering an end-to-end trainable framework with considerable practical value. While the contribution is strong, there are some areas where minor refinements could enhance the overall quality of the manuscript:

• The usage of f versus F throughout the equations needs clarification to ensure readers can follow the theoretical development without confusion.
• The authors should provide a more thorough justification for their choice of α=2 for quadratic entropy, as this parameter selection significantly impacts the theoretical foundation of their approach.
• Including a brief analysis of computational complexity would help readers understand the practical implications and scalability of implementing this framework.

Author Response

See attached pdf.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

1. The unit of the horizontal coordinate in Figure 9 is inconsistent with the caption. 

2. In Line 261 and Eq. (13), the descriptions of f_L and f_l are inconsistet. 

3. Equation (31), line 531: Mixed usage of norm notation (∥·∥² vs. ∥·∥₂) appears throughout the manuscript. Please standardize to ∥·∥₂ for L2 norm representation to maintain mathematical rigor and consistency with conventional notation.

Author Response

See attached pdf

Author Response File: Author Response.pdf