| Title |
Experimental Study on the Expansion of the Short-term Prediction Range of Rebar Prices Using Deep Learning |
| Authors |
이용성(Lee, Yong-Seong) ; 김경환(Kim, Kyung-Hwan) |
| DOI |
https://doi.org/10.5659/JAIK.2020.36.12.265 |
| Keywords |
Deep learning; Stack bidirectional LSTM; Prediction range expansion; Rebar price |
| Abstract |
This study presents a method for expanding the prediction range of rebar price prediction using the short-term prediction method of deep
learning. In general, the prediction range of a short-term prediction is dependent on the time interval of the data to be entered, so it can be
expanded by adjusting the time interval of the data. However, as the range of forecasts increases, the size of the data decreases, which can
lead to overfitting that cannot guarantee good results. The average accuracy of the forecasts is approximately 98.49% when the scope of the
forecasts is extended from 1 month to 2 and 3 months with the proposed approach presented in this study. In addition, this approach could
be used as a basis for expanding the predictive range of deep learning in a study that predicts prices with time series data including
common building materials. |