농식품 소비 트렌드 기반 작목 선택 예측 모형

Crop Selection Prediction Model for Agricultural and Food Consumption Trend

This study develops a crop selection prediction model using agricultural and food consumption trend. Under the expected utility hypothesis (EUH), we devise a crop selection mechanism where farmers choose crops providing the maximum expected utility from their profit. Then, feed forward neural network (FFNN) model is used for the prediction of the expected utility, and vector autoregressive (VAR) model is constructed with atypical Big Data based on web-based text query concerning consumption trend. We apply models to the exclusive consumption trend panel data provided by the Rural Development Administration (RDA) for six specific regions. Empirical results show that various crops are predicted to provide the maximum expected utility and that the prediction model can be improved by considering atypical Big Data.