Dataview:
list from [[]] and !outgoing([[]])📗 -> Regularization
🎤 Vocab
L2 or Ridge Regularization - The L2 penalty aims to minimize the squared magnitude of the weights
- Pneumonic could be the 2 in L2 stands for squared
L1 or Lasso - The L1 penalty aims to minimize the absolute value of the weights
- Pneumonic could be that the 1 means the power of 1 is penalized
Hyperparameters:
Hyper-parameters are “higher-level” parameters that describe structural information about a model that must be decided before fitting model parameters, examples of hyper-parameters we discussed so far:
- Learning rate alpha , Regularization lambda.
❗ Information
Regularization is a method to discourage overfitting
Difference between L1 and L2:
L2 shrinks all coefficients, but does not eliminate them
L1 can eliminate them (shrink to zero). It performs feature selection
📄 -> Methodology
✒️ -> Usage
- How and where is it used
🧪-> Example
- Define examples where it can be used
🔗 -> Links
Resources
- Put useful links here
Connections
- Link all related words