In machine learning, parameters are variables that are learned from the training data. These parameters enable the model to make predictions or classifications on new, unseen data. The specific parameters depend on the type of machine learning algorithm. Here are examples.
Linear Regression:
- Intercept (bias): A constant term added to the linear equation.
- Coefficients (weights): Weights assigned to each input feature.
Logistic Regression:
- Intercept (bias): A constant term added to the logistic equation.
- Coefficients (weights): Weights assigned to each input feature.
Support Vector Machines (SVM):
- Kernel parameters: These define the kernel function.
- Regularization: How to maximize margin v. minimize errors.
Decision Trees:
- Split criteria: How to split data at each node in the tree.
- Maximum depth: Controls complexity and potential for overfitting.
Random Forest:
- Number of trees: The number of individual decision trees.
- Maximum depth: The maximum depth of each decision tree.
Neural Networks:
- Weights and biases: Parameters that define neural connections.
- Learning rate: Controls the step size during optimization.
K-Nearest Neighbors (KNN):
- Number of neighbors (K): How many neighbors make a prediction.
- Distance metric: How to measure distance between data points.