This paper presented the application of the Stochastic Gradient Boosting Tree-based approach for finding the solubility of ILs using different measurable parameters. The predicted values were compared with the values obtained from the different well-known model and experimental studies. Furthermore, the obtained results were compared with other soft computing methods such as Genetic expression programming and LSSVM models. Although the work is not novel but demonstrates the application of SGB tree for calculating the solubility of the ILs. But authors made all possible efforts to qualify for the work for publication. Hence, the reviewer suggests incorporation of the following comments enhance the quality and make it more qualified for possible publication. The author is suggested to correct sentence structures and some randomly mistake in the paper. The structure of the paper can be modified, such as in results and discussion many equations are mentioned. The reviewer thinks that this is not the proper place to discuss it. Hence another heading can be added to discuss them. Also the working platform and coding language etc. was not mentioned in the paper, which is important to discuss and mentioned so, that the research can be fully utilized for the ease of mankind and other research applications. Limitation and advantage of SGB method over other methodologies can be discussed. The model can be a complex one for the ILs or for any other application using SGB mythology, there should be some mentioning so that application can be fully utilized in the industrial or research work. Rest small comments are mentioned below for further correction and necessary action to incorporate in the article. • Page 3, para1 line 3: degree sign missing. • Page 10, para1 line 1: acentric factor/coefficient abbreviation is missing. • In Table 1, data do not seem to be aligned. • In the last column of Table 2, an abbreviation for the parameter is missing. • In Fig 3: the x-axis shows Data Number, kindly correct this typo error. • IN Fig 7, 14 and 17: x-axis and y-axis labels are hiding the values of the x and y parameters, it can be placed at some proper place. • In Fig 1 author chosen 1000 number of trees as optimum, but how this value was obtained, even 500 values seems to be okay as a stopping criterion for preventing overfitting. • On page 14, Table 3 is mentioned for SGB tree model with R and other Error models, but actually, it shows tuning parameters for SGB model. Which is already mentioned for another purpose. Therefore, it is suggested to correct it. • It is difficult to understand to the reviewer that, how the critical values for the ILs were utilized in the model, kindly explain it.