| |
RSC_BURGES reduced set construction - burges method
|
a = rsc_burges(alg,hyper)
generates a rsc object, using the burges method
hyperparameters:
child=svm algorithm worked on
rsv=.5 if rsv e [0,1): #rsv = #sv * rsv
if rsv > 1: #rsv
max_it=100 maximum of iterations for gradient descent
eps=1e-5 epsilon
phase2=yes perform burges phase2 ?
minimizer=1 (1) gradient descent (0) conj. gradient
stv=3 number of restarts
deriv='der_rbf' use function, which name is supplied as string, as
function to calculate d||w-w*||^2/dz_l
if set to 'num', usee numerical derivative instead
model:
alpha new alphas for rs-vectors
Xsv rs vectors
stats:
w2=0 final value of ||w-w*||^2
delta_w=0 total decrease in ||w-w*||^2
dw=0 run through delta_w
methods:
train constructs a reduced set, returns trained rs-machine
test tests new rs-machine on supplied data
example:
d=gen(toy2d('2circles','l=100'));
[r,a]=train(svm({kernel('rbf',1),'C=10000','alpha_cutoff=1e-2'}),d);
[r,a2]=train(rsc_burges(a,'rsv=.9'),d);
test(a2,d,loss);
author: chris burges
reference: simplified sv decision rules, 1996