Data Space
Is the coordinate-space where data samples are mapped to.
In a classic machine-learning system we have samples of data supplied as vectors with each field in the vector corresponding to some dimension in data-space. For example a sample set of flowers with the fields:
- Petal-length
- Leaf-color
- Height
Will be supplied as a set of 3 length vectors with \( X_0 \) being the petal-length for sample \( X \), \( X_1 \) being the leaf-color, etc.
If we plot these samples directly in some euclidean system we get the data-space.