“…Other methods have also been considered, such as multiplicative regression Zhao et al, 2014;Kepaptsoglou et al, 2017) [3,22,23], two-stage least square regression (2SLS) (Taylor et al, 2004;Estupiñán and Rodriguez, 2008) [24,25], Poisson regression [3,6], negative binomial regression [11], and structural equation modelling (SEM) (Sohn and Shim, 2010) [8]; geographical methods such as distance-decay weighted regression [1] and the network Kriging method (Zhang and Wang, 2014) [26]; machine learning methods such as the decision tree (DT) and support vector regression (SVR); and item-based collaborative filtering methods based on cosine similarity (CF) (Hu et al, 2016) [27], cluster analysis (Deng and Xu, 2015) [28], and back propagation neural networks (BPNN) (Li et al, 2016) [29]. However, understanding the results is a major challenge in terms of the interpretability of the function modeled by the machine learning algorithm.…”