gelnet (version 1.2.1)

gelnet.oneclass: GELnet for one-class regression

Description

Constructs a GELnet model for one-class regression using the Newton method.

Usage

gelnet.oneclass(X, l1, l2, d = rep(1, p), P = diag(p), m = rep(0, p),
  max.iter = 100, eps = 1e-05, w.init = rep(0, p), silent = FALSE)

Arguments

X

n-by-p matrix of n samples in p dimensions

l1

coefficient for the L1-norm penalty

l2

coefficient for the L2-norm penalty

d

p-by-1 vector of feature weights

P

p-by-p feature association penalty matrix

m

p-by-1 vector of translation coefficients

max.iter

maximum number of iterations

eps

convergence precision

w.init

initial parameter estimate for the weights

silent

set to TRUE to suppress run-time output to stdout (default: FALSE)

Value

• A list with one element: [object Object]

Details

The function optimizes the following objective: $$-\frac{1}{n} \sum_i s_i - \log( 1 + \exp(s_i) ) + R(w)$$ where $$s_i = w^T x_i$$ $$R(w) = \lambda_1 \sum_j d_j |w_j| + \frac{\lambda_2}{2} (w-m)^T P (w-m)$$ The method operates by constructing iteratively re-weighted least squares approximations of the log-likelihood loss function and then calling the linear regression routine to solve those approximations. The least squares approximations are obtained via the Taylor series expansion about the current parameter estimates.

See Also

gelnet.lin, gelnet.oneclass.obj