A Comparative Study of Machine Learning Models with Hyperparameter Optimization Algorithm for Mapping Mineral Prospectivity
Round 1
Reviewer 1 Report
Thank you for revising this article however a number of points remain and, now the manuscript is more clear, I have identified some further questions I hope you will address.
I hope these comments are helpful, please find them in the attached file.
Best wishes,
Comments for author File: 
Comments.pdf
Author Response
Please see the attachment
Author Response File: Author Response.pdf
Reviewer 2 Report
The answer to Point 6: Equation 6 is wrong
Response 6: We have modified the symbolic representation in Equation 12 (original Equation 6).
Equation 12 appears to be a Gaussian kernel (Hayton, P., SchÓ§lkopf, B., Tarassenko, L., & Anuzis, P. (2000). Support vector novelty detection applied to jet engine vibration spectra. In T. K. Leen, T. G. Dietterich & V. Tresp (Eds.), Proceedings of the 13th Conference on Advances in Neural Information Processing Systems (NIPS’ 2000) (pp. 946–952). Breckenridge, USA: MIT Press.) as Eq.3 in Ref.47 (Chen, Y.; Wu, W. Mapping mineral prospectivity by using one-class support vector machine to identify multivariate geological anomalies from digital geological survey data. Australian Journal of Earth Sciences. 2017, 64(5), 639-651. [doi:10.1080/08120099.2017.1328705])
K(x,y)= < Fi(x),Fi(y) >_H=exp(-||x-y||^2/2sigma^2)
But you have
K(x,y)=Fi(x),Fi(y)_H=exp( - x – y^2 /2sigma^2)
It seems that < > and || || were lost.
Author Response
Please see the attachment
Author Response File: Author Response.pdf
Reviewer 3 Report
The paper is written well with proper analyses. Only it needs more discussion for future work and the cons of the methodologies. Where more research can be done to improve the approaches?
Author Response
Please see the attachment
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
I thank the authors for addressing much of my previous review and resolving Points 1,2,3,5,6,8,10,12. A few minor points exist that I hope you consider as I believe they will make the study more comparable to other mineral potential modelling studies.
Point 4 – thank you for an eloquent explanation that is clear. If this can be incorporated into the manuscript, I think it would be very beneficial to the reader.
Point 7 – Thank you for your additional explanations in the manuscript. I still find this aspect confusing and do not understand why a classification problem that investigates class probabilities is not implemented. Regardless of different models, one should be trying to model and report the same target variable for fair comparison – I refer back to my previous review comments regarding the conventional way to report mineral prospectivity models. I do not wish to suggest we enforce convention at the detriment to innovation but the current models and presentation make it very difficult to offer fair comparison within this study and to other similar studies of mineral potential.
Point 9 – Again, this is perhaps a point of convention. TPR and TNR are not the conventional methods for plotting a ROC curve and (as I understand it) it should be plotted as Sensitivity (TPR) vs 1-Specificity (1-TNR). This may be my pedantry creeping in, but for comparison to other studies this could cause a point of confusion.
Point 11 – Thank you for including these. A minor point of clarification, “setting to default” is fine, but some reference to what algorithm is being used needs to be made e.g. Python’s scikit-learn module for example. What is default in one application/coding language may be completely different in another.
Best wishes,
Author Response
Please see the attachment.
Author Response File: 
Author Response.docx
This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.
Round 1
Reviewer 1 Report
I have attached my review in a word document for you to refer to. I hope you find these comments useful.
Kindest regards,
Comments for author File: Comments.pdf
Reviewer 2 Report
Ref 2-3 should be merged – it seems that ref. 2 has been split into two lines and followed Ref numeration shifted
References for the machine learning algorithms: supervised learning [1], unsupervised learning, [2,3], semi-supervised learning [4-6], reinforcement learning [7], artificial neural network (ANN) [8,9], decision tree [10], adaptive neural fuzzy system [11], constrained boltzmann machine [12], support vector machine (SVM) [13-16], random forest (RF) [17,18], extreme learning machine (ELM) [19], MaxEnt model [20], genetic algorithm (GA) [21,22], particle swarm optimization (PSO) [23-25], bat algorithm (BA) [26], firefly algorithm (FA) [27], ant colony algorithm (ACA) [28] should be on the original papers where they were first introduced or on the some more common reviewing sourse.
Line 87: “The output steps of MLP are as follows [34]” – the Ref [34]( Raghu, S.; Sriraam, N. Optimal configuration of multilayer perceptron neural network classifier for recognition of intracranial epileptic seizures. Expert Systems with Application. 2017, 89, 205-221.) does not describe the followed procedure.
Line 106: ”processes of AdaBoost algorithm are as follows [38,39]” - again, one could not find the followed steps in these Refs (Tharwat, A.; Gaber, T.; Hassanien, A.E.; Elhoseny, M. Automated toxicity test model based on a bio-inspired technique and AdaBoost classifier. Computers and Electrical Engineering. 2018, 71, 346-358., Wang, L.; Lv, S.X.; Zeng, Y.R. Effective sparse adaboost method with ESN and FOA for industrial electricity consumption forecasting in China. Energy. 2018, 155, 1013-1031.)
Line 127: v and m were not defined.
Equation 6 is wrong
Subsection 3.2.2: It seems that ETM+, “ZY-1” 02C and ASTER could be described
Equation 13: yd=se – sp is not common used definition of Youden index (yd=se+sp-1), why such a formula was used in this work? It needs to be more detailed explained. Seems that in this work the value 1-specify taken as a specify.
Line 316: missprint – Aadboost
The term AUC occurs many times much earlier than it was defined in section 4.6.
