A predictive-analytics problem, adapted from a PuzzlOR puzzle: 20 historical patients are recorded with their symptoms (sneezing, sore throat, fatigue, stuffy nose, etc.) and their confirmed diagnosis (Cold, Flu, or Allergies).
Question: Patient #21 hasn’t been diagnosed yet, but is exhibiting a stuffy nose, sneezing, and a sore throat. Using only the data from the 20 prior patients, rank the three possible diagnoses (Cold, Flu, Allergies) by how likely Patient #21 has each.
The interesting part isn’t just getting an answer — it’s representing this kind of predictive/probabilistic reasoning in a decision model in a way that a subject-matter expert (not just a programmer or statistician) can follow.
Send your solutions to DecisionManagementCommunity@gmail.com, or open a pull request to add yours here.