Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This work is about the field of reservoir computing, offering actionable insights for parameter tuning. The paper would be suitable for publication with minor revisions to address the below points:

1.The paper assumes fixed values for other system parameters (e.g., injection strength κinj​, feedback strength κ). A brief discussion on how these might interact with θ and τ could provide a more holistic view.

2.The robustness of the identified optimal parameters under varying conditions (e.g., noise, parameter drift) should be included to strengthen the practical relevance.

3.The manuscript would benefit from a comparison with other state-of-the-art RC systems or alternative approaches to highlight the novelty and competitiveness of the proposed optimizations.

Author Response

Dear Reviewer,

Thank you very much for your comments concerning our manuscript entitled “Time parameters optimization for the semiconductor laser-based time-delay reservoir computing system” (Manuscript ID: photonics- 3593230). Your comments are all valuable and very helpful for revising and improving this paper, as well as the important guiding significance to our researches. We have studied these comments carefully and have tried our best to make according modifications which are highlighted in the revised manuscript. The detailed responses to reviewer comments are listed as follows.

Responses to Reviewer 1 Comments:

This work is about the field of reservoir computing, offering actionable insights for parameter tuning. The paper would be suitable for publication with minor revisions to address the below points:

1. The paper assumes fixed values for other system parameters (e.g., injection strength κ_inj, feedback strength κ). A brief discussion on how these might interact with θ and τ could provide a more holistic view.

Response:

We acknowledge you very much for your professional advice. In the third paragraph of section “3.3 Discussion”, some sentences (lines 318-326) were added to address this issue.

2. The robustness of the identified optimal parameters under varying conditions (e.g., noise, parameter drift) should be included to strengthen the practical relevance.

Response:

We acknowledge you very much for your professional advice. In the section “3.3 Discussion”, Figure 9 and the corresponding description (lines 299-314) were added to illustrate the system’s robustness.

3. The manuscript would benefit from a comparison with other state-of-the-art RC systems or alternative approaches to highlight the novelty and competitiveness of the proposed optimizations.

We acknowledge you very much for your kind reminders. In the fourth paragraph of section “3.3 Discussion”, some sentences (lines 341-353) and references were added to clarify this issue.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Dear Editor,

About the Manuscript ID photonics-3593230

Title: Time parameters optimization for the semiconductor laser based time-delay reservoir computing system

Authors: Qiudi Li , Yushuang Hou , Keqiang Li , Chunxia Hu , Dianzuo Yue

            This work focuses on the field the time-delay reservoir computing systems, with the those based of semiconductor lasers. Using the Santa Fe time series prediction task and memory capacity evaluation task; the authors investigate the influence of key time parameters virtual node interval, delay feedback, and data injection period on the performance of SL-based time-delay RC systems. The results are encouraging and can provide valuable insights for optimizing time-delay RC systems, enabling better task-specific performance and stability.

            The subject is worth of investigation and fits well the scope of the Journal of MDPI Photonics.  I have fully gone through the manuscript and appreciate the efforts undertaken by the authors on the impact of time parameters on the performance of SL-based time-delay RC systems. I think that this manuscript merit publication, but before that, several points need to be addressed, and some explanations need to be added.

1) The introduction part should be improved by incorporating recent review concerned the time-delayed reservoir computing using mutually coupled multimode semiconductor laser for high-speed image recognition, see for example:

https://doi.org/10.1016/j.optlastec.2025.112774

2) The authors should ameliorate the clarities of the figures.

3) Figure 4, need more explanation. How the minimum prediction error has been determined?

4) Especially, this study provides only experimental results and discussion, is there any theoretical work in this context to confirm the experimental data?

5) In Fig 7 for the MCmax which varies with θ under τ = T. What is the signification of the peak at 0.5 (θ/Tro),?

6) The authors should indicate the limit of their method of study. What is the disadvantage of using the this method?

7) Instead of general conclusion, the authors need to optimize their findings for the technology specific where their findings could be employed along with merits for the same over the existing device architecture.

Comments on the Quality of English Language

English need minor correction.

Author Response

Dear Reviewer,

Thank you very much for your comments concerning our manuscript entitled “Time parameters optimization for the semiconductor laser-based time-delay reservoir computing system” (Manuscript ID: photonics- 3593230). Your comments are all valuable and very helpful for revising and improving this paper, as well as the important guiding significance to our researches. We have studied these comments carefully and have tried our best to make according modifications which are highlighted in the revised manuscript. The detailed responses to reviewer comments are listed as follows.

Responses to Reviewer 2 Comments:

This work focuses on the field the time-delay reservoir computing systems, with those based of semiconductor lasers. Using the Santa Fe time series prediction task and memory capacity evaluation task; the authors investigate the influence of key time parameters virtual node interval, delay feedback, and data injection period on the performance of SL-based time-delay RC systems. The results are encouraging and can provide valuable insights for optimizing time-delay RC systems, enabling better task-specific performance and stability.

The subject is worth of investigation and fits well the scope of the Journal of MDPI Photonics.  I have fully gone through the manuscript and appreciate the efforts undertaken by the authors on the impact of time parameters on the performance of SL-based time-delay RC systems. I think that this manuscript merit publication, but before that, several points need to be addressed, and some explanations need to be added.

1. The introduction part should be improved by incorporating recent review concerned the time-delayed reservoir computing using mutually coupled multimode semiconductor laser for high-speed image recognition, see for example: https://doi.org/10.1016/j.optlastec.2025.112774.

We acknowledge you very much for your kind reminders. The progress in RC based on multimode SLs is described at the end of the second paragraph in the “Introduction” (lines 70-73), and a reference [26] is added.

2. The authors should ameliorate the clarities of the figures.

We acknowledge you very much for your kind reminders. We have enhanced the output resolution of exported images.

3. Figure 4, need more explanation. How the minimum prediction error has been determined?

We regret any ambiguity in our initial presentation. Additional explanatory sentences (lines 204-208) have been incorporated into the description of Fig. 4 to address this matter.

4. Especially, this study provides only experimental results and discussion, is there any theoretical work in this context to confirm the experimental data?

We sincerely appreciate the reviewer for providing this valuable research direction from a professional perspective. Indeed, theoretical analysis is crucial for validating experimental data. However, in this work, we have focused solely on simulation studies, with experimental validation conducted on the Santa Fe time series prediction task and memory capacity evaluation task. Due to time constraints for revision, we were unable to complete the theoretical analysis of this issue, and we hope for the reviewer’s understanding. In future work, we plan to first implement the system in hardware to verify our simulation results, followed by in-depth theoretical studies to investigate the underlying physical mechanisms. We have added a statement (lines 373-375) addressing this matter at the end of the “Conclusion” section.

5. In Fig 7 for the MC_max which varies with θ under τ = T. What is the signification of the peak at 0.5 (θ/T_ro)?

We acknowledge you very much for your professional advice. We have added several sentences (lines 270-275) in the description of Fig. 7 to explain this issue.

 

6. The authors should indicate the limit of their method of study. What is the disadvantage of using the method?

We acknowledge you very much for your professional advice. In the third paragraph of section “3.3 Discussion” some sentences (lines 318-332) were added to explain this issue.

7. Instead of general conclusion, the authors need to optimize their findings for the technology specific where their findings could be employed along with merits for the same over the existing device architecture.

We acknowledge you very much for your kind reminders. In the third paragraph of section “3.3 Discussion” and in the second paragraph of “Conclusion”, some sentences (lines 332-340, lines 364-373) are added to explain this issue.

We hope that the revised manuscript will meet with approval.

Once again, thank you very much for your professional comments and suggestions.

Yours sincerely,

Qiudi Li, Yushuagn Hou, Keqiang Li, Xiaoyu Guo, Chunxia Hu, Dianzuo Yue

Author Response File: Author Response.pdf