Feature Extraction

Overview

Feature extraction is the process of taking a dataset and computing features derived from the values in the dataset. Sometimes feature extraction is necessary because the learning algorithm being used is linear, for example, but the patterns in the data is not linear. Feature extraction may be able to create linear patterns by simply transforming the features in the dataset.

Topics

The following are some common feature extraction techniques.

• Time Series Feature Extraction
• Normalization : is the process of scaling all the values of a dataset. Typically this means scaling the numeric value from its measured value to the number of standard deviations it is away from the mean of the dataset at hand.
• Dimensions - a common method of feature extraction is to reduce the number of the dimensions in the data.
• Kernel Methods are a set of methods that can convert an algorithm that recognizes linear patterns to one that can recognize non-linear patterns. Example algorithms that can be extended using kernel methods:
  
  
  • perceptron
  • linear regression
  • principal component analysis
  • clustering