Experiments in machine learning with Go Run
go run main.go
to see it in action. It will generate a 1000x5 matrix A
using a uniform random distribution,
a 5x1 matrix x
using a random normal distribution, and a matrix y = Ax + e
, where e
is
a 1000x1 error matrix (normally distributed with std. deviation = 0.1. The library then solves
the equation Ax=y for x using the
Moore-Penrose pseudoinverse.
It does ok! It is pretty quick!
Contains a matrix struct that knows how to do some linear algebra
Implements linear regression.