Huber norm approximation to the tilted absolute value cost function

Share:

Description

Huber norm approximation to the tilted absolute value cost function used to fit a QRNN model. Optional left censoring is supported.

Usage

1
2
qrnn.cost(weights, x, y, n.hidden, tau, lower, eps, Th,
          Th.prime, penalty)

Arguments

weights

weight vector of length returned by qrnn.initialize.

x

covariate matrix with number of rows equal to the number of samples and number of columns equal to the number of variables.

y

predictand column matrix with number of rows equal to the number of samples.

n.hidden

number of hidden nodes in the QRNN model.

tau

desired tau-quantile.

lower

left censoring point.

eps

epsilon value used in huber and related functions.

Th

hidden layer transfer function; use sigmoid for a nonlinear model and linear for a linear model.

Th.prime

derivative of the hidden layer transfer function Th.

penalty

weight penalty for weight decay regularization.

Value

numeric value indicating tilted absolute value cost function, along with attribute containing vector with gradient information.

See Also

qrnn.nlm, qrnn.fit