linearly separable

Terms from Artificial Intelligence: humans at the heart of algorithms

Two sets of data are linearly separable if there is a straight line (or hyperplane in higher dimensions) that separates them. In other words there is a linear combination (ai) of the features and a threshold T such that fp
      a1d1 + a2d2 + ... + andn is less than T for data items (di) in the first datasets and greater than T for the second
In such cases, the hyperplane acts as a {[classifier}}.

Used on pages 111, 135

Also known as linear separability