Optimasi Manajemen Kapasitas Pelayanan Pasien Rawat Jalan dengan Pendekatan Machine Learning

Abstract

The main problem of this research is the maximum capacity that is not optimal because not all outpatient requests are fulfilled, the limited range of appointment times for returning patients and the average appointment lead time for old patients due to emergency conditions so that new patients are unpredictable. The study focuses on optimizing the previous model by restating the objective of this system in terms of the risk of violation (ε). The prerequisites are the potential for not being able to fulfill every request from a patient and not exceeding. The model's goal is to determine the necessary minimal capacity for a particular breach risk in order to achieve the intended lead time. This model fixes the constraints of the prior model by determining the equivalent deterministic of individual opportunity constraints, which will depend on what is known about the probability distribution, and optimizes the management of outpatient requests that are not all fulfilled, the limited range of appointment times for returning patients, and the average appointment lead time for old patients. The test results show that the probability value of the number of first visits (FV) patients arriving at time unit i and being given an appointment at time unit j is greater than or equal to the stochastic arrival of FV patients in time unit i is 0.51. Each test is different because it involves generating random values from computer memory. The probability value has increased from the previous model.