Developer Terms and Conditions › The Development › Molecular Sensing Models › Trying to get a True/False element answer from a model. › Reply To: Trying to get a True/False element answer from a model.
January 13, 2016 at 2:29 pm #2549
Generally speaking, a properly sized collection to provide meaningful results or analysis should include no less than 40 different samples and have 3 scans per sample (or more or less depending on your samples and your needs).
Can you share what is the desired model you want to create? Would you like to classify 3 different pills?
In this case, less than 40 samples is enough.
The Consumer Physics Team