A Method of Using a Scoring Algorithm to Find Similar Diagnoses in Medical Information Systems

This paper considers the solution of the classification problem, which is one of the data mining problems based on the databases of medical information systems. The general mathematical formulation of the ranking problem is given, the existing algorithms such as RankBoost, RankSVM and IR-SVM are analyzed, and their applications are mentioned. At the same time, the advantages and disadvantages of solving the disturbance problem using a properly distributed neural network model of a neural network calculated by heuristic methods are noted. In addition, the six-step scoring algorithm developed for solving the classification problem was used to find similar diagnoses by solving the irritation problem using the first three steps, and the results were obtained. The obtained results were verified based on the information of the medical information system database, and the reliability of the algorithm was compared. A block diagram of an adapted version of the scoring algorithm for solving the ranking problem is presented.