|A doctor examines a female patient. (Photo credit: Wikipedia)
Recently I was asked to find a solution to the following problem:
Dr Who is a popular doctor with lots of experience in treating a particular cosmetic condition. Dr Who has multiple clinics in different geographic locations. How do we scale up Dr Who’s experience to all the clinics?
This was my answer:
Dr Who uses 2 machines A & B and 2 products C & D.
Dr Who uses machine A with two settings A1 and A2 and B with 2 settings B1 and B2.
So you have several combinations and permutations of machine settings and products.
Dr Who extracts 3 information from a patient, X, Y, and Z. X and Y are captured by a machine and Z is noted by him.
Dr Who has 10000 patients with X,Y,Z known. He has data on what combination worked for whom and what did not.
We can correlate values of X,Y,Z for each patient with the combination (A1/A2/B1/B2/C/D) that worked for him. So when a patient comes to us, we measure X,Y,Z and propose the most likely combination to work for him. So our assessment will be based on Dr Who’s experience on 10000 patients.