Abstract: The accuracy and efficiency of optimization algorithms remain an essential challenge in supervised machine learning. We used gradient descent and Newton's method, along with backtracking ...