데이터마이닝을 위한 다항식 기반의 유전적 프로그래밍 기법과 조선분야 응용

초록

Engineering data contains experts’ experience and know-how in its own. It is very useful to extract knowledge from the accumulated existing data by using data mining technique. This can be utilized as a tool of knowledge creation in knowledge management system. This paper deals with generating optimal polynomials using genetic programming. Low orders Taylor series are used to make the polynomial easily approximate nonlinear function or domain data. The suggested method uses the minimal or very small size of learning data. This method can be used in engineering problem, especially ship and ocean engineering domain that has insufficient existing data to perform machine learning.