Effective Heart Disease Prediction Using Machine Learning Techniques

Round 1

Reviewer 1 Report

1. Why DBSCAN for detecting outliers?

2. For dataset, specify as CVD dataset, instead of saying as generic - dataset

3. Table 4, best values can be made bold

4. Work flow model diagram can be added.

Author Response

Please find in the attached file the reviewer responses.

Author Response File: Author Response.pdf

Reviewer 2 Report

 

Paper algorithms-2155347 “Effective Heart Disease Prediction using Machine Learning Techniques”

 

Comments

This study focuses on effective heart disease prediction using machine learning techniques. I think the paper fits well the scope of the journal and addresses an important subject. However, a number of revisions are required before the paper can be considered for publication. There are some weak points that have to be strengthened. Below please find more specific comments:

 

*The abstract seems to be detailed and thorough. No comments.

*Keywords: The keywords seem to be adequate. Please another one or two relevant keywords.

*The introduction section could benefit from more statistical information to better highlight the importance of the main subject at hand.

*The literature review seems to be reasonable for the most part. Please double check for the most recent and relevant studies published over the last 2-3 years. The entire literature review section could be more detailed.

*Section 3 seems to be adequate.

*Please provide more details regarding the input data used throughout the experiments. Some supporting references would be helpful to justify the data selection.

*The manuscript contains quite a few figures and tables. Please double check and try to provide a more detailed description of these figures and tables where appropriate to make sure that the future readers will have a reasonable understanding of what these figures and tables represent.

*The results section should be more detailed and thorough. The outcomes of this research and evaluation of the proposed methodology should be much more detailed.

*Future research: This study primarily used machine learning methods for heart disease prediction. In the future research, the authors should explore more advanced optimization algorithms for this decision problem. Therefore, the authors should create a general discussion regarding the importance of advanced optimization algorithms (e.g., customized heuristics and metaheuristics) for challenging decision problems. There are many different domains where advanced optimization algorithms have been applied as solution approaches, such as online learning, scheduling, multi-objective optimization, transportation, medicine, data classification, and others (not just the decision problem addressed in this study). The authors should create a discussion that highlights the effectiveness of advanced optimization algorithms in the aforementioned domains and their potential applications for the decision problem addressed in this study. In the future research, the proposed machine learning methods could be actually compared to advanced heuristic and metaheuristic algorithms. This discussion should be supported by the relevant references, including but not limited to the following:

An online-learning-based evolutionary many-objective algorithm. Information Sciences 2020, 509, pp.1-21.

A diploid evolutionary algorithm for sustainable truck scheduling at a cross-docking facility. Sustainability 2018, 10(5), p.1333.

Ambulance routing in disaster response considering variable patient condition: NSGA-II and MOPSO algorithms. Journal of Industrial & Management Optimization 2022, 18(2), p.1035.

Such a discussion will help improving the last section of the manuscript significantly.

Author Response

Please find attached reviewer responses.

Author Response File: Author Response.pdf

Reviewer 3 Report

In this paper, the authors tackled an important problem of automated prediction of heart diseases using machine learning. The topic is certainly worthy of investigation, but the manuscript suffers from quite a number of important shortcomings which should be, in my opinion, thoroughly addressed before the paper could be considered for publication:

1.       The authors should clearly specify what kind of data modality they are targeting in this work – this should be clearly presented in the abstract.

2.       The authors should revisit the literature on predicting cardiovascular diseases and cardiovascular events using machine learning – there are quite a number of new techniques available in the literature and validated over clinical data (see e.g., the work by Drozdz: https://cardiab.biomedcentral.com/articles/10.1186/s12933-022-01672-9). The authors should indeed discuss such approaches in detail to better contextualize their work within the state of the art, and to better highlight the contribution behind the work reported here.

3.       Please revise Table 1 – there are some unnecessary characters here and there (e.g., “-Kaggle Dataset”) and missing spaces. Also, were the methods investigated using the same dataset splits for the Kaggle set (training/test)? If not, then comparing those algorithms is inherently biased and cannot be considered reliable.

4.       Please make sure that all abbreviations are defined at their first use.

5.       Figure 1 is actually a table, please rework it (also, its quality is very poor).

6.       Please provide more detailed quantitative characteristics of all parameters from the investigated dataset.

7.       There are lots of vague statements across the manuscript such as “variables appear to have some outliers” – please quantitatively prove that there are indeed outliers (please present the distributions of the parameters).

8.       The captions of the figures should be placed below them.

9.       How did the authors split the dataset into 80/20 training/test? Was it a random split or was it delivered by the authors of the dataset (also, please provide a direct link to the dataset).

10.   The number of digits after the decimal point in all numbers in the correlation table should be consistent.

11.   Please improve the formatting of Table 4.

12.   Please provide a link to the repository containing the implementation of the proposed method in order to ensure its full reproducibility.

Author Response

Please find attached reviewer response.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Second Review

 

The authors have adequately addressed most of my comments. One comment has not been fully addressed. The authors did include a more detailed discussion regarding the evaluation of the proposed method against the alternative solution approaches as a part of the future research. However, this discussion has not been supported by any references. In order to make this discussion more convincing, please provide some supporting references for the future readers, as suggested in my original report.

Author Response

Thank you for the comment. References have been added at relevant places.

Reviewer 3 Report

Thank you indeed for addressing my concerns.

Author Response

Thank you for your feedback.