Threshold-Feedforward Neural Network (T-FNN) for Tractor Automated Ground Leveling (AGL)

This paper first introduces the concept of using the Supervised Machine Learning method to improve Tractor Automated Ground Leveling (AGL) operation. This paper further explores various supervised machine learning that included Feedforward Neural Network (FNN), Learning Vector Quantization Neural Network (LVQNN), Naïve Bayes (NB), and Support Vector Machine (SVM). The results show that Threshold-Feedforward Neural Network (T-FNN), a modified version of FNN is the most effective method among all algorithms for applications such as AGL that have no clear separation data.