Predicting Chain’s Manufacturing SME Credit Risk in Supply Chain Finance Based on Machine Learning Methods

Round 1

Reviewer 1 Report

This paper uses machine learning methods to predict the manufacturing SME's credit risk. However, the proposed risk evaluation indicators are lack of theoretical foundation. The authors don't give the the reasons why they choose the four machine learning methods. Are these methods applicable? Is there any other more effective method? There are no answers in this paper. Overall, the contribution of this paper is not enough. 

Author Response

Dear peer reviewer

     

Thank you for your thoughtful comments, which have helped significantly improve this paper. We appreciate your hard work for reviewing our paper and hope this revision is satisfactory. Please see the attachment for our specific reply to your comments on this article.

Thank you so much

Yours sincerely,

Author Response File: Author Response.pdf

Reviewer 2 Report

According to the manuscript Ref. No. sustainability-1965268 it is observed that:
1. the authors made a preliminary selection of the financial credit risk evaluation indicators for the supply chain of manufacturing SMEs.
2. The idea is interesting and impotant.
3. The methodology used are correct and all the obtained results are interpreted correctly.

Author Response

Dear peer reviewer

     

Thank you for your thoughtful comments, which have helped significantly improve this paper. We appreciate your hard work for reviewing our paper and hope this revision is satisfactory. Please see the attachment for our specific reply to your comments on this article.

Thank you so much

Yours sincerely,

Author Response File: Author Response.pdf

Reviewer 3 Report

Supply chain finance is an effective way to solve the financing problems of small and medium-sized manufacturing enterprises, and the assessment of credit risk is one of the key issues of manufacturing supply chain financing. To my knowledge, imbalanced data  classification problems is one of the challenging task for predicting chain’s manufacturing SME credit risk in supply chain finance based on machine learning methods.

1. The ratio of different classes should be provided for training and testing datasets.

2. The k-fold cross-validation procedure provides a good general estimate of model performance that is not too optimistically biased, at least compared to a single train-test split.

3. Accuracy is not the metric to use when working with an imbalanced dataset. I recommend looking at the following performance measures that can give more insight into the accuracy of the model than traditional classification accuracy: Confusion Matrix, Precision, Recall, F1 Score (or F-score), Kappa (or Cohen’s kappa) and ROC Curves.

4. Contributions and novelties should be strengthen by comparing your work with previous studies.

5. Datasets are too small. How do you ensure obtain the convincing conclusion?

Author Response

Dear peer reviewer

     

Thank you for your thoughtful comments, which have helped significantly improve this paper. We appreciate your hard work for reviewing our paper and hope this revision is satisfactory. Please see the attachment for our specific reply to your comments on this article.

Thank you so much

Yours sincerely,

Author Response File: Author Response.pdf

Reviewer 4 Report

The study addresses an important topic. Generally, studies focus on the risk analysis of companies in isolation and do not address supply chain risk.

The authors make use of an adequate theoretical framework.

On line 127, there are 5 references [14,23,24,25,26]. I suggest reporting what each author addressed in their study.

I recommend creating a session presenting the research method, separating it from the other sessions.

Further detail the form of data collection. How did you collect the data? Only documental analysis or through questionnaire or interview?

Author Response

Dear peer reviewer

 

Thank you for your thoughtful comments, which have helped significantly improve this paper. We appreciate your hard work for reviewing our paper and hope this revision is satisfactory. Please see the attachment for our specific reply to your comments on this article.

Thank you so much

Yours sincerely,

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The authors have addressed all my concerns. This version of manuscript has been sufficiently improved to warrant publication in Sustainability.

Reviewer 3 Report

Accept