Models & Architectures · Level 2 of 5
Gradient Boosting
Building an additive model by repeatedly fitting learners to improve an objective.
Later learners address the current model's remaining errors or gradients.
Example
Small trees are added sequentially to improve predictions.
Listen to the definition and example
Audio transcript
Gradient Boosting. Building an additive model by repeatedly fitting learners to improve an objective. Later learners address the current model's remaining errors or gradients. For example: Small trees are added sequentially to improve predictions.
Explore this concept
Why it matters
This helps you distinguish a model’s structural assumptions from its learned parameter values.
Related concepts
Quick recall question
Try answering before looking back at the definition.