“…Machine learning models have become widespread in a variety of applications related to geoscience and natural disaster assessments. The most widely used machine learning models in the gully susceptibility domain (called shallow machine learning models) are support vector machines (e.g., Pourghasemi et al, 2017; Rahmati et al, 2017), artificial neural networks (ANNs) (e.g., Roy et al, 2020; Shahabi et al, 2019), random forest (e.g., Pal et al, 2020; Pham et al, 2020; Saha et al, 2020), maximum entropy (e.g., Azareh et al, 2019), boosted regression trees (e.g., Amiri et al, 2019; Hembram et al, 2021; Wang et al, 2021), flexible discriminant analysis (e.g., Gayen et al, 2019), quick, unbiased, efficient statistical tree (e.g., Soleimanpour et al, 2021), randomized tree (e.g., Wang et al, 2022), naïve Bayes (e.g., Garosi et al, 2019; Hitouri et al, 2022; Lana et al, 2022), best‐first decision tree (e.g., Lei et al, 2020), least absolute shrinkage and selection operator (e.g., Pourghasemi et al, 2020), extreme gradient boosting (e.g., Chen et al, 2021), and extreme gradient boosting (e.g., Chen et al, 2021). Furthermore, artificial intelligence/machine learning models have been widely used for different geomorphic processes, including landslide, soil erosion, and land subsidence (e.g., Marjanović et al, 2011; Merghadi et al, 2020; Mohammady et al, 2019; Sahour et al, 2021; Sekkeravani et al, 2022).…”