Model Based vs. Instance Based Learning
Model Based vs. Instance Based Learning
- In model based learning, models learn weights/underlying distribution from the sample dataset
- I.E. Neural Network
- In instance based learning, model use the whole dataset as features
- I.E. K-nearest neighbour
Model-based methods (e.g. Neural Network) learn parameters; instance-based methods (k-nearest Neighbor (KNN)) keep the training data as the model.