gdls: Least squares with graident descent
In cmna: Computational Methods for Numerical Analysis

Description Usage Arguments Details Value See Also Examples

Solve least squares with graident descent

1	gdls(A, b, alpha = 0.05, tol = 1e-06, m = 1e+05)

`A`	a square matrix representing the coefficients of a linear system
`b`	a vector representing the right-hand side of the linear system
`alpha`	the learning rate
`tol`	the expected error tolerance
`m`	the maximum number of iterations

gdls solves a linear system using gradient descent.

the modified matrix

Other linear: choleskymatrix(), detmatrix(), invmatrix(), iterativematrix, lumatrix(), refmatrix(), rowops, tridiagmatrix(), vecnorm()

1
2
3

head(b <- iris$Sepal.Length)
head(A <- matrix(cbind(1, iris$Sepal.Width, iris$Petal.Length, iris$Petal.Width), ncol = 4))
gdls(A, b, alpha = 0.05, m = 10000)