Machine learning and neural network based model predictions of soybean export shares from US Gulf to China

Abstract

In this paper, we propose a general model for the soybean export market share dynamics and provide several theoretical analyses related to a special case of the general model. We implement machine and neural network algorithms to train, analyze, and predict US Gulf soybean market shares (target variable) to China using weekly time series data consisting of several features between January 6, 2012 and January 3, 2020. Overall, the results indicate that US Gulf soybean market shares to China are volatile and can be effectively explained (predicted) using a set of logical input variables. Some of the variables, including shipments due at US Gulf port in 10 days, cost of transporting soybean shipments via barge at Mid-Mississippi, and soybean exports loaded at US Gulf port in the past 7 days, and binary variables have shown significant influence in predicting soybean market shares.

Hypothesis tests on high dimensional datastreams, with finance application

“…Here, we present the results of the predictions using various machine-learning techniques [25]. We first use logistic regression (LR).…”

Financial Decisions Based on Zero-Sum Games: New Conceptual and Mathematical Outcomes

“…In this paper, P is a linear functional (scalar or inner product) and zero-sum games with incomplete information (see De Meyer et al (2010)) are based on it. On the other hand, linear functionals are used in machine learning too, where it is known that artificial neural networks are models influenced by the structure and function of biological neural networks in animal brains (see Awasthi et al (2022)). In particular, scalar products underlie a classification algorithm that makes its estimations combining a set of weights with the feature vector.…”