Understanding the Gradient Boosting Machine algorithm
Gradient Boosting Machine (GBM) is a forward learning ensemble ML algorithm that works on both classification as well as regression. The GBM model is an ensemble model just like the DRF algorithm in the sense that the GBM model, as a whole, is a combination of multiple weak learner models whose results are aggregated and presented as a GBM prediction. GBM works similarly to DRF in that it consists of multiple decision trees that are built in a sequence that sequentially minimizes the error.
GBM can be used to predict continuous numerical values, as well as to classify data. If GBM is used to predict continuous numerical values, we say that we are using GBM for regression. If we are using GBM to classify data, then we say we are using GBM for classification.
The GBM algorithm has a foundation on decision trees, just like DRF. However, how the decision trees are built is different compared to DRF.
Let’s try to understand...
