Dimension Reduction can be achieved via 2 methods:
 |
Feature Selection: The main aim of doing feature selection is to reduce the search space (for attributes), by selecting relevant features and then removing the remaining irrelevant features. That is, feature selection selects "m" features from the entire set of "n" features such that m ≤ n. Ideally m <<< n. |
|
|
Feature Transformation: It is also known as PCA, or Principle Component Analysis. Feature transformation forms a new set of feature from the original set, but does not actually reduce the original dimensionality. One of it's many uses is for visualization of the data. |