Gradient-Descent Boosted Training System

A Gradient-Descent Boosted Training System is an boosted ensemble supervised learning system that is a gradient-descent system that can solve a gradient descent boosted training task).

• Context:
  • It can range from being a Gradient Boosting Estimation Ssytem to being a Gradient Boosting Ranking System to being a Gradient Boosting Classification System.
  • It can (often) be a Gradient Boosting-based Decision Tree Learning System.
  • …
• Example(s):
  • Gradient Boosting-based Relational Learning System, such as Natarajan et al., 2012).
  • CatBoost.
  • XGBoost.
  • LightGBM.
  • …
• Counter-Example(s):
  • Bagging-based Training System.
  • Random Forest System.
  • AdaBoost System
  • Simple Decision Tree Learning System, such as C4.5.
• See: Iterative Gradient Descent; Ensemble Learning, Boosting Meta-Algorithm, Differentiable Function.

---

---

---