See Model fit

See Model fit metrics

Gradient boosting is a machine learning ensemble method that sequentially combines weak learners to create a powerful predictive model by minimizing errors using gradient descent.

Gradient Boosting Additive Models (GBAM) is a variant of gradient boosting that extends the traditional algorithm to handle additive models. An additive model is a model that consists of a sum of functions of individual input variables, where each function depends only on a single input variable.

Gradient boosting minimizes a specified loss function by using gradient descent optimization. It computes the gradients of the loss function and adjusts the subsequent weak learner to correct the errors made by previous learners. The learning process aims to iteratively reduce the loss and improve the overall model performance.