Combination of K-Means and Profile Matching for Drag Substitution

Abstract

Substitution of medicines can be done when the medicine required by a patient in the pharmacy is exhausted. In substituting prescription medicines, the pharmacist must notify patients that medicine substitution contains same active ingredients in equal amounts, same type, and its price is different. The right of medicine substitution belongs to pharmacists as long as the doctor allows the substitution. This method serves to make some similar medicines into a particular data set. This researched used K-means method. In terms of time, this method can be used to reduce the searching time process of medicines required by patients. Selection profile matching method was used to determine the medicine substitute recommendations based on the most ideal medicines selection for patients. The recommended medicine’s substances were exactly the same as the previous medicines listed in the prescription. Evaluation clustering used elbow method with cluster k = 3. Based on the experiments, the recommendation of medicine substitution with K-means and profile matching produced accuracy was 0.93506.