Skip to content
Ilya Gyrdymov edited this page Sep 18, 2017 · 7 revisions

Vector in algebraic sense is a sequence of numerical elements. There are several operations which can be applied to this sequence. The operations are:

  • Vector addition (both operands are vectors, result of the operation - new vector)
  • Vector subtraction (both operands are vector, result of the operation - new vector)
  • Scalar multiplication (one of operands is a scalar and another is a vector, result of the operation - new vector)
  • Scalar product of vectors (aka dot product, both operands are vectors, result of the operation - a scalar)
  • Addition of a scalar and a vector
  • Subtraction a scalar from a vector

Vector is an atom of machine learning and, of course, is a base structure of this library. A couple of classes of the structure is implemented in simd_vector library.

Why is vectors so important for machine learning? Answer is quite simple: because it is very convinient to union features of a considered object into such a structure.