Gradient Descent - shivamvats/notes GitHub Wiki

  • Conjugate Gradient Descent: It can solve Ax = b where A is symmetric and positive-definite. It is the go-to method to solve such systems when A is large and sparse. For dense matrices, go for decomposition.