Predicting Health Risks of Adult Asthmatics Susceptible to Indoor Air Quality Using Improved Logistic and Quantile Regression Models
, Shayma Alkobaisi 2,*, Matthew Horak 3, Choon-Sik Park 4, Sungroul Kim 5 and Joel Davidson 1Round 1
Reviewer 1 Report
This manuscript is among the most clearly written ones I’ve reviewed in the past few years. As an air quality specialist, I am not qualified to comment on the machine learning algorithms used to estimate population and individual exposure risks. But the predictive modeling framework makes full sense to me. I would like to see more applications of the same or similar framework to exposure assessment, given the increasing deployment of wearable monitoring devices.
Author Response
The attached file includes authors' responses to reviewer's comments.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
In the paper “Predicting health risks of adult asthmatics susceptible to
indoor air quality using improved logistic and quantile
regression models”, the authors present a new approach to predict the risks of patients with asthma using a new model. The paper is novel, interesting and battles an important aspect in epidemiology. Moreover, the paper is clearly and well-written. However, there are certain issues that should be addressed
When building a new statistical predictive model to determine-predict health risks, this model should undergo specific tests, using statistical tests. Specifically, in every new model the authors should
-compare the data fitting of the new versus the old model, we (using−2 log likelihood statistics)
-calibration of the model (Hosmer-Lemeshow test)
-discrimination ability (using receiver operating characteristic (ROC) curve)
-internal validation
-external validation
Author Response
The attached file includes authors' responses to the reviewer's comments.
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
the authors covered all my concerns
