Please enable JavaScript.
Coggle requires JavaScript to display documents.
Decision Tree, image - Coggle Diagram
Decision Tree
Pros
Interpretability
Mixed Data Type
Non linear relationships
Understanding Feature importance
Handling missing values
Quick prototyping
Reduce Overfitting
Bagging=bootstrap + Aggregate
Random Forest- Extension of Bagging
Random Sampling Training Data
Random Subset of Input Features
Bootstrap
Aggregation
Majority rule: Add all results and take majority
Prunning
Pre-pruning
Limit depth
Require minimum impurity before split
Post-pruning
Grow tree to full size, then prune
Cons
Overfitting
Lack of robustness-sensitive to small changes in input data
Entropy: to score or measure impurity
