DEVICE AND METHOD FOR ESTIMATING PRICE OF POLY VINYL CHLORIDE RELATED MATERIALS

The present application relates to a device and method for estimating the price of a polyvinyl chloride related material using a convolutional neural network and long-term memory. The device for estimating the price of a polyvinyl chloride related material includes: a first learning data input unit which receives, as first learning data, a polyvinyl chloride resin index, a related material price, customs clearance data and turn-around work related data, for a certain period of time, selected as a learning target; a first learning unit which extracts a feature map indicating the direction of price fluctuations of polyvinyl chloride; a first characteristic extraction unit which extracts the characteristic of the direction of the price fluctuations of polyvinyl chloride; a second characteristic extraction unit which extracts the characteristic of an absolute price of a polyvinyl chloride related material; a second learning unit which receives each characteristic extracted by the first characteristic extraction unit and the second characteristic extraction unit as second learning data, and performs deep learning on the same; and an output unit which reflects the result of the second learning unit, derives an estimated price of the polyvinyl chloride related material, and outputs the derived estimated price of the polyvinyl chloride related material. 본 출원은 컨볼루션 신경망 및 장단기 메모리를 이용한 폴리염화비닐 관련재 가격 예측 장치 및 방법에 관한 것이다.