A Novel Model Selection Framework for Forecasting Agricultural Commodity Prices using Time Series Features and Forecast Horizons

Abstract

The oscillations of agricultural commodity prices have abundant impact on people's daily lives and also the inputs and outputs of agricultural production. To take proper decisions one should require an accurate forecast of commodity prices. Accuracy of crop price forecasting techniques is important because it enables the supply chain planners and government bodies to take appropriate actions by estimating market factors such as demand and supply. In emerging economies such as India, the crop prices at marketplaces are manually entered every day, which can be prone to human-induced errors like the entry of incorrect data or entry of no data for many days. In addition to such human prone errors, the fluctuations in the prices itself make the creation of stable and robust forecasting solution a challenging task. To forecast prices more adaptively, this study proposes a novel model selection framework which includes time series features and forecast horizons. Twenty-nine features are used to depict agricultural commodity prices and three intelligent models are specified as the candidate forecast models; namely, artificial neural network (ANN), support vector regression (SVR), and extreme learning machine (ELM). Both random forest (RF) and support vector machine (SVM) are applied to learn the underlying relationships between the features and the performances of the candidate models. Additionally, a minimum redundancy and maximum relevance approach (MRMR) is employed to reduce feature redundancy and further improve the forecast accuracy. The trial that's what results exhibit, firstly, the proposed model determination system has a superior figure execution contrasted and the ideal competitor model and basic model normal; besides, highlight decrease is a useful way to deal with further work on the exhibition of the model determination structure; and thirdly, for bean and pig grain items, various disseminations of the time series highlights lead to an alternate determination of the ideal models.