Multitask Learning Based on Least Squares Support Vector Regression for Stock Forecast
Round 1
Reviewer 1 Report
Actual topic which fits into special issue. Strong mathematical background, good experimental part. Related works are enough, including works from last three years. Results are interesting.
You should add discussion section, where you will compare your results with results obtained from related works.
Minor comments:
- line 492 ... k=8 and N=9 ... - you should add space around =
- line 495 ... alpha=0.05 is ... - you should add space around =
Author Response
Response to Reviewer 1 Comments
Major comments:
Point 1: You should add discussion section, where you will compare your results with results obtained from related works.
Response 1: Thanks for your constructive comment. We reorganized the content of our manuscript, and added the discussion section to compare the experimental results of the proposed models with those of the other four algorithms. In the discussion section, the prediction performance of our proposed multitask learning models was evaluated from a statistical testing perspective. Besides, we also analyzed the advantages of the proposed models and elaborated the future work directions from a fair level. All the revised contents are updated in the revision manuscript, which are highlighted with gray.
Minor comments:
Point 1: - line 492 ... k=8 and N=9 ... - you should add space around =.
Response 1: Thanks for your warm suggestion. We carefully checked the original manuscript and revised the formula formatting in line 492. All the revised contents are updated in the revision manuscript, which are highlighted with gray.
Point 2: - line 495 ... alpha=0.05 is ... - you should add space around =.
Response 2: Thanks for your warm comment. We carefully checked our manuscript and revised the formula formatting in line 495. At the same moment, we carefully revised all the formula format problems in the revision manuscript. All the revised contents are updated in the revision manuscript, which are highlighted with gray.
Author Response File: Author Response.pdf
Reviewer 2 Report
Thank you for the paper. The authors have to improve provided version:
1. The formating of equation 8 has to be revised
2. The name of section 4 has to be changed into "Experiments"
3. The authors has to place the discussion section before the conclussions.
4. Under discussion part, the authors have to show what theoretical or methodological gap the paper fills.
5. Conclussions section has to be extended. The authors have to present the limitation of their research.
6. The authors have to present most significant results of their empirical research under conclussions section.
7. The sentence about future research directions, has to be moved to the discussion section.
8. The list of references has to be extended. The authors have to include contemprorary papers from 2022.
9. The quality of Figure 2 has to be improved.
10. The authors has to explain the abbrevations used in formulas. For example, in equation (1) the authors have to explain J and T meaning. The same is valid for other formulas. The authors have to check carefully and provide the meanings for the readers of their paper accordingly.
Author Response
Response to Reviewer 2 Comments
Major comments:
Point 1: The formatting of equation 8 has to be revised.
Response 1: Thanks for your warm suggestion. We carefully checked our manuscript and revised the formatting of Equation (8). Meanwhile, we carefully checked all the formula format problems in the revision manuscript. All the revised contents are updated in the revision manuscript, which are highlighted with gray.
Point 2: The name of section 4 has to be changed into "Experiments".
Response 2: Thanks for your warm suggestion. We reorganized the content of our manuscript, and renamed the section 4 as “Experiments” in the revision manuscript. All the revised contents are updated in the revision manuscript, which are highlighted with gray.
Point 3: The authors has to place the discussion section before the conclusions.
Response 3: We are grateful for your constructive comment. We rearranged the content of the original manuscript, and added the discussion section to perform a comparison for the experimental results of the proposed models with those of the other four algorithms. In the discussion section, the prediction performance of our proposed multitask learning models was evaluated from a statistical testing perspective. Besides, we also analyzed the advantages of the proposed models and elaborated the future work directions from a fair level. All the revised contents are updated in the revision manuscript, which are highlighted with gray.
Point 4: Under discussion part, the authors have to show what theoretical or methodological gap the paper fills.
Response 4: Thanks for your warm suggestion. We reorganized the content of our manuscript and added the discussion section in section 5. The prediction ability and robust performance of the proposed multitask learning models was evaluated from a statistical testing perspective. The advantages of EMTL-LS-SVRs are mainly benefited from they can effectively capture the correlation among multiple tasks and flexibly select the appropriate kernel functions for the shared information and private information respectively. All the revised contents are updated in the revision manuscript, which are highlighted with gray.
Point 5: Conclusions section has to be extended. The authors have to present the limitation of their research.
Response 5: We appreciate for your valuable suggestion. We rewrote the conclusions section of the original manuscript of the proposed models and mainly extended the limitations and the future work directions in detail. All the revised contents are updated in the revision manuscript, which are highlighted with gray.
Point 6: The authors have to present most significant results of their empirical research under conclusions section.
Response 6: Thanks for your warm suggestion. We rewrote the conclusions section of the original manuscript, and perform a supplement on the empirical results carefully. Meanwhile, the limitations and future work directions are elaborated in detail. All the revised contents are updated in the revision manuscript, which are highlighted with gray.
Point 7: The sentence about future research directions, has to be moved to the discussion section.
Response 7: Thanks for your warm suggestion. We rearranged the content of our manuscript and added the discussion section. In section 5, we mainly analyzed the advantages of the proposed models and elaborated the future work directions from a fair level. All the revised contents are updated in the revision manuscript, which are highlighted with gray.
Point 8: The list of references has to be extended. The authors have to include contemporary papers from 2022.
Response 8: Thanks for your valuable comment. We have reorganized the references list in the revision manuscript and extended some latest literature, such as the literature [4], [9] and [32]. All the revised contents are updated in the revision manuscript, which are highlighted with gray.
Point 9: The quality of Figure 2 has to be improved.
Response 9: Thanks for your warm suggestion. We revised Figure 2 in the revision manuscript, and carefully checked all the figures appearing in this paper. All the revised contents are updated in the revision manuscript, which are highlighted with gray.
Point 10: The authors has to explain the abbreviations used in formulas. For example, in equation (1) the authors have to explain J and T meaning. The same is valid for other formulas. The authors have to check carefully and provide the meanings for the readers of their paper accordingly.
Response 10: We are grateful for your constructive comment. We carefully checked all the formula in the original manuscript and explained the corresponding abbreviations in formulas. At the same moment, we carefully revised all the formula format problems in the revision manuscript. All the revised contents are updated in the revision manuscript, which are highlighted with gray.
Author Response File: Author Response.pdf