Paper Title
Machine Learning for Diagnosis of Acute Abdominal Pain in Adults at Suratthani Hospital

Abstract
Machine learning, a branch of Artificial Intelligent, plays vast variety roles in our everyday life. We applied machine learning in medical application for diagnosis of diseases that cause patents an acute abdominal pain. We aim to provide a useful guide for an appropriate decision making in medical management. The investigation was based on a clinical data based from Suratthani Hospital during September-December 2017 with 407 cases, 15 diagnostic parameters and 13 diagnoses. Random Forest, Naïve Bayes, Support Vector Machine, Nearest Neighbors, Logistics Regression were trained with 308 cases and separately tested on 99 cases. No major differences in overall accuracy were observed (53-60%), except for support vector machine that was lowest (57%). Keywords - Acute Abdominal Pain, Machine Learning, Suratthani Hospital, Diagnostic Accuracy, Acute Appendicitis.