Why is it called SVR?

• This method is called SVR because each point is seen as a vector. Every point outside the tube is a support vector, which will dictate the formation of the tube

$\epsilon$ Insensitive Tube

• One of the easy approach to think about SVR is it is a linear regression but instead of a line it is a tube (so considerably for room for data point discrepancy to the actual line)
• This is the Ordinary Least Square Sum formula for the tube and epsilon

The support vector are all the points (vector) outside the tube because those point in a way shape and structure the tube and will affect the formula for the equation to find the minimal error (hence this method is call support vector regression)