Linear Regressing using Tensor Flow
Overview
Tensor Flow can be used to find the solution to a
linear regression
problem.
For many linear regression problems, the method of
least squares
is used because it produces an exact analytic solution that does not require and optimization engine such as tensor-flow.
Using an optimization engine, one is free to change the
error function
used to grade the output to something that may not have an exact analytic solution, however, the engine will require some time to
execute and will not return an exact solution.
Topics
