Description
Quadratic and Linear Programming
solves quadratic programming problem for pattern recognition for support vectors
solve the quadratic programming problem
minimize c’ x + 1/2 x’ H x
subject to Ax = b
l <= x <= u
for a documentation see R. Vanderbei, LOQO: an Interior Point Code
for Quadratic Programming
n : number of primal variables
m : number of constraints (typically 1)
h_x : dot product matrix (n.n)
a : constraint matrix (n.m)
b : constant term (m)
l : lower bound (n)
u : upper bound (m)
primal : workspace for primal variables, has to be of size 3 n
x = primal; n
g = x + n; n
t = g + n; n
dual : workspace for dual variables, has to be of size m + 2 n
y = dual; m
z = y + m; n
s = z + n; n
verb : verbosity level
sigfig_max : number of significant digits
counter_max: stopping criterion
restart : 1 if restart desired
Design a complete device with the initial interface for different sensors using Wolfram Mathematica
Reviews
There are no reviews yet.