Time Series Forecasting of Agricultural Products’ Sales Volumes Based on Seasonal Long Short-Term Memory

Tae-Woong Yoo,Il-Seok Oh

doi:10.3390/app10228169

Abstract

In this paper, we propose seasonal long short-term memory (SLSTM), which is a method for predicting the sales of agricultural products, to stabilize supply and demand. The SLSTM model is trained using the seasonality attributes of week, month, and quarter as additional inputs to historical time-series data. The seasonality attributes are entered into the SLSTM network model individually or in combination. The performance of the proposed SLSTM model was compared with those of auto_arima, Prophet, and a standard LSTM in terms of three performance metrics (mean absolute error (MAE), root mean squared error (RMSE), and normalization mean absolute error (NMAE)). The experimental results show that the error rate of the proposed SLSTM model is significantly lower than those of other classical methods.

Highlights

Agricultural commodity prices are volatile, and they are affected by various factors, including weather conditions, the planting area, consumption trends, and policy factors [1]
We propose an agricultural sales forecasting system based on deep learning to stabilize agricultural supply and demand
In the long short-term short-term memory memory (LSTM) and seasonal long short-term memory (SLSTM) models, the sales volume datasets were split into three subsets, with the first 60% of the data considered as the training set, the 20% as the validation set, and the most recent 20% as the test set

Summary

Introduction

Agricultural commodity prices are volatile, and they are affected by various factors, including weather conditions, the planting area, consumption trends, and policy factors [1]. Accurate forecasting of agricultural supply and demand is essential for both agricultural producers and consumers. The trends of the agricultural market are complicated and nonlinear as a result of numerous factors, including economic globalization, financial speculation, climate change, and oil price fluctuations. This complexity makes it increasingly difficult to accurately predict agricultural supply and demand [1,5]. Since it is very difficult to predict agricultural supply and demand with different variability factors, more research needs to be conducted on such predictions. Studies on energy and electrical load predictions, which have less volatile factors than agricultural predictions, are actively conducted. The studies on other applications include rainfall forecasting [12], bitcoin price forecasting [13], bankruptcy prediction [14], and stock price prediction [15,16]

Methods

Results

Discussion

Conclusion