Abstract
This study deals with a development of a past similar cases reference system for supporting engineer's decision making based on the Neural Network learning and Self Organization Map. The system shows similar cases as a map of two dimensions constructed by Self Organization Map to engineer, and weight parameters of features are updated by case selected by an engineer based on Neural Network learning algorithm. This paper describes a concept of the past similar cases reference system and details of the developed system configuration. To estimate the system performance and efficiency of our idea, results about applied 16 sample case data are indicated.
