Questions
What is the name of the operation which approximates a continuous function by a linear combination of basis functions?
What kind of traits do they have? Methods of function expansion mentions orthogonal bases/kernels from which I suppose each basis would be a orthogonal vector in a vector space. However, I am not sure if polynomial functions are orthogonal vectors.
Fourier transfer uses trigonometric functions $sin(\theta), cos(\theta)$.
Bayesian regression uses gaussian.
Gaussian process looks using covariance.
Polynomial regression uses polynomial.
