Feature Maps (
In machine learning, pattern recognition and image processing, a feature map starts from an initial set of measured data and builds derived values (also known as features) intended to be informative and non-redundant, facilitating the subsequent learning and generalization steps, and in some cases leading to better human interpretations.
A feature map is related to dimensionality reduction; it involves reducing the amount of resources required to describe a large set of data. When performing analysis of complex data, one of the major problems stems from the number of variables involved. Analysis with a large number of variables generally requires a large amount of memory and computation power, and may even cause a classification algorithm to overfit to training samples and generalize poorly to new samples.
When the input data to an algorithm is too large to be processed and is suspected to be redundant (for example, the same measurement is provided in both pounds and kilograms), then it can be transformed into a reduced set of features, named a feature vector.
The process of determining a subset of the initial features is called feature selection. The selected features are expected to contain the relevant information from the input data, so that the desired task can be performed by using the reduced representation instead of the complete initial data.