Naive Bayes
Train Naive Bayes to predict categorical outcomes
Probabilistic classifier based on Bayes' theorem with "naive" independence assumption.
When to use:
- Text classification
- Need fast training and prediction
- Want probabilistic predictions
- Small to medium datasets
Strengths: Very fast, works with small data, naturally handles multiclass, probabilistic Weaknesses: Assumes feature independence (rarely true), less accurate than modern methods
Model Parameters
Var Smoothing (default: 1e-9) Portion of the largest variance of all features added to variances for stability.