A Machine Learning Method for the Quantitative Detection of Adulterated Meat Using a MOS-Based E-Nose
Round 1
Reviewer 1 Report
The article titled "A Machine Learning Method for the Quantitative Detection of Adulterated Meat Using a MOS-Based E-Nose" deals with Machine learning methods applied to a sensor array. The article deals with a very complex problem that is difficult to solve effectively with the analysis of VOCs alone. I believe that the data treated here are very poor but translated with advanced machine learning methods we try to make sense.
Some notes that might be helpful:
- I recommend doing even lower%
- do you measure the same sample for 10 days? so is it a shelf-life analysis? of meat that is no longer edible? have you considered doing microbiological analyzes of the sample?
- how were the 10 sensors chosen? based on what characteristics? what compounds do you think can be generated after 10 days of analyzing a sample (biogenic amines?)
- how long does the measurement last in total? tiga 115 to 127
- at what T ° are the samples kept during sampling?
- figure 2 does not make sense, ask to add: A) the same sensor as it varies at different%. B) to make all the measurements between sampling and recovery including the base line. (which R / R0?)
- it seems that there are several sensors that do not give any useful contribution, that is they are not able to discriminate why they are used?
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Dear Authors,
A PDF file with review is attached.
Regards
Comments for author File: Comments.pdf
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
