A MACHINE-LEARNING-BASED MODEL FOR FORECASTING MEDICAL DEVICE FOREIGN TRADE

Abstract

Forecasting the medical device foreign trade is a very important issue and a challenging problem due to many external artifacts in the medical device market for making an efficient policy. Many reports, including the simple statistical based methods do not provide sufficient forecasting for foreign trade and this problem may be solved using a machine-learning based approach. The purpose of this study is to introduce an efficient model for forecasting medical device foreign trade. In this respect, export and import data obtained with 54 different commodity codes were performed using some machine-learning algorithms. The best prediction performance was achieved with SVM regression model with the average R2=0.974 and for the last five years. In 2025, total medical device exports and imports are expected to be $1.03 billion and $2.12 billion, respectively. We also performed Market Penetration Index and Product Diversification Index to analyze medical device foreign trade.