The concept have a United States and other creating nations, the pace of care unit (CU) use has been expanding massively over the previous decade. Because of the maturing populace and the ever-developing interest for restorative consideration, viable administration of patient, progress among various consideration offices will demonstrate the need of shortening the length of clinic remain, improving patient results, dispensing basic consideration assets and diminishing preventable readmission. An epic system was created by treating an arrangement of change occasions as a point procedure, for demonstrating the patient move through different CUs and mutually anticipating patients' goal CUs and span days. An epic discriminative learning calculation was proposed targeting improving the forecast of change occasions on account of inadequate information as opposed to learning a generative point process model through greatest probability estimation. The proposed model was parameterized as a commonly adjusting procedure, the estimation issue was figured through summed up straight models which fit productive learning dependent on exchanging bearing technique for multipliers (ADMM).