All reviews of published articles are made public. This includes manuscript files, peer review comments, author rebuttals and revised materials. Note: This was optional for articles submitted before 13 February 2023.
Peer reviewers are encouraged (but not required) to provide their names to the authors when submitting their peer review. If they agree to provide their name, then their personal profile page will reflect a public acknowledgment that they performed a review (even if the article is rejected). If the article is accepted, then reviewers who provided their name will be associated with the article itself.
The Authors have addressed all of the reviewers' comments, therefore it is now ready for publication.
[# PeerJ Staff Note - this decision was reviewed and approved by Xiangjie Kong, a PeerJ Section Editor covering this Section #]
no comment
The authors significantly updated the experimental design, by focusing on the research direction 2 suggested by the reviewer.
no comment
The author significantly enhanced the manuscript, by removing redundant research topics, and focusing on the "Survival analysis in breast cancer". The authors revised all sections of the manuscript to focus on one research topic. By focusing on the primary topic, the author enhanced the novelty and significance of the revised paper.
All comments have been added in detail to the 4th section called additional comments.
All comments have been added in detail to the 4th section called additional comments.
All comments have been added in detail to the 4th section called additional comments.
Review Report for PeerJ Computer Science
(Survival analysis in breast cancer: Evaluating ensemble learning techniques for prediction)
Thanks for the revision. I examined the responses to the reviewer comments and the final version of the paper in detail. Although some answers are limited, they are sufficient. Due to the contribution of the study to the literature and the improvements made, I recommend that this research paper be accepted. I wish the author success in future works. Best regards.
The most important issues in the current manuscript involve identifying a clearer title, providing complete references, fixing figure captions, and addressing all comments related to model comparisons. Finally, I invite the Authors to fix the spelling and formatting issues.
Therefore, based on the previously highlighted points, the Manuscript requires some Major revisions.
**PeerJ Staff Note:** Please ensure that all review and editorial comments are addressed in a response letter and that any edits or clarifications mentioned in the letter are also inserted into the revised manuscript where appropriate.
**Language Note:** The review process has identified that the English language must be improved. PeerJ can provide language editing services - please contact us at [email protected] for pricing (be sure to provide your manuscript number and title). Alternatively, you should make your own arrangements to improve the language quality and provide details in your response letter. – PeerJ Staff
The paper contains numerous grammatical errors. I recommend that the authors consider utilizing a language editing service for assistance. Although there are many more, a few suggestions are given below:
Line 63: CHANGE to the model this complex data TO model this complex data.
Line 70: CHANGE million life worldwide TO million lives worldwide
Line 72: CHANGE will shed lights TO will shed light
Line 77: CHANGE method analysing TO method for analysing
Line 79: CHANGE time takes from TO time it takes from
Line 96: CHANGE variety fields TO variety of fields
Line 107: CHANGE RSF the second TO RSF is the second
Line 150: CHANGE evaluation made through TO evaluation was made through
Line 152: CHANGE that the both methods TO that both methods
Line 157: CHANGE comparison the proposed TO comparison of the proposed
Line 163: CHANGE datasets in employed TO dates employed
Line 167: CHANGE method is enable TO method is to enable
Line 176: CHANGE those which utilized TO those utilized
Line 241: CHANGE 2509 observation TO 2509 observations
Line 241: CHANGE of them removed from TO of them were removed from
Line 246: CHANGE same of three subjects TO same for three subjects
Line 279: CHANGE to grown the TO to grow the
Line 360: CHANGE if the more TO if more
Line 481: CHANGE affect the how TO affect how
Line 482: CHANGE methods performances TO methods perform
Line 540: CHANGE models have TO models yield
Line 541: CHANGE results and so that the TO results thereby fulfilling the
Line 559: CHANGE to better understanding TO in gaining a better understanding
Line 565: CHANGE prevention the progress TO prevention of the progress
No Comment
No Comment
Comment 1: The title should be revised. Since you are using different Ensemble Learning models for the prediction of Breast Cancer, I suggest the title should be more descriptive about it. A few suggestions are given below:
"Improving Breast Cancer Survival Prognosis with Ensemble Learning Algorithms"
Survival Analysis in Breast Cancer: Evaluating Ensemble Learning Techniques for Prediction"
Enhanced Prediction of Breast Cancer Progression Using Advanced Survival Models"
"Comparative Analysis of Survival Prediction Models for Breast Cancer Progression: A Study Using Ensemble Learning Techniques"
Comment 2: Make sure all the references have a DOI. If you use Mendeley you can search a DOI lookup option for that.
Comment 3: On page 44, the caption of Figure 6 is incomplete.
The manuscript "Ensemble Learning Models Can Outperform Traditional Cox Model in Breast Cancer Survival" evaluated the Cox Proportional Hazards (PH) model random Survival Forest (RSF) and Conditional Inference Forest (Cforest) model with the event time's data in the Machine Learning task of breast cancer progression prediction. The result shows that the Cox PH model has a lower C-index and bigger prediction error than the RSF and the Cforest approach.
The manuscript is a typical manuscript in the machine learning (ML) research field to propose and evaluate a better ML solution to solve an existing prediction problem.
However, the authors should focus on one primary topic. However, the title and abstract may confuse the readers: 1) Ensemble learning models can outperform traditional Cox model, and the authors evaluate and compare these models with breast cancer survival data; 2) The prediction of breast cancer survival is essential, but it is hard to do that, the authors present that Ensemble learning models can outperform traditional Cox model in predicting the breast cancer survival. So, the reviewer recommends the authors update the abstract to start with the importance of breast cancer survival prediction rather than starting with "event times data are …"
Research direction 1: the comparison of RSF, Cforest, and Cox in analyzing time-event data
If the authors want to propose that the RSF and Cforest models are better than the Cox model in handling event-time data, they should make a more systematic comparison of these models rather than evaluating them with one dataset (e.g., the breast cancer survival data). The authors can keep the current manuscript structure but need to add more evaluation datasets, not just the breast cancer datasets.
Research direction 2: the comparison of RSF, Cforest, and Cox in predicting breast cancer survival
Suppose the authors want to propose that the RSF and Cforest models are better than the Cox model in predicting breast cancer survival. Then, the author can keep using the current evaluation dataset but needs to rewrite most of the manuscript's content to focus on the breast cancer prediction topic.
The authors should pay more attention to the format/structure of the manuscript, for example, Line 218 (redundant new lines), Line 235 (need to add a new line), Line 286 (need to re-organize the locations of equations)
All comments have been added in detail to the 4th section called additional comments.
All comments have been added in detail to the 4th section called additional comments.
All comments have been added in detail to the 4th section called additional comments.
Review Report for PeerJ Computer Science
(Ensemble learning models can outperform traditional Cox model in breast cancer survival)
1. Within the scope of the study, in addition to the Cox Proportional Hazards model for breast cancer progression prediction, Random Survival Forest and Conditional Inference Forest were used for machine learning-based ensemble learning models.
2. The introduction and related works sections need to be detailed. The difference of the study from the literature and its main contributions to the literature should be stated more clearly and in bullet points. It is also recommended to add a table regarding the literature in columns such as the model used, the problem addressed, the dataset used, results, shortcomings, plus points, etc.
3. Within the scope of the study, two different datasets available in the literature were preferred. The reason for choosing these datasets should be explained more clearly. The fact that a data set specific to the study was not used limits the originality of the study in terms of data set.
4. Although there are many different machine learning approaches in the literature that can be used to predict breast cancer progression, only Random Survival Forest and Conditional Inference Forest ensemble models were used in this study in addition to the Cox Proportional Hazards model. What is the reason for choosing these two models in particular? Apart from this, have there been experiments with different machine learning models?
5. The point of originality in terms of the model should be stated more clearly. Both the limited models used and the use of existing models in the literature greatly limit the originality of the study in terms of model.
6. Although the results obtained depending on the models used in the study are acceptable, there are big question marks regarding the depth of the study.
As a result, although the study is of a certain quality, the steps mentioned above must be fully answered and/or explained.
All text and materials provided via this peer-review history page are made available under a Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.