costMatrix

integer vector of positive values (>=1)
  representing the labels of the instances

either a cost matrix to check for consistency
  with labels in y, or a character string defining the
  standard matrix to compute.  If a character string the
  accepted values are "0/1" for a 0-1 cost matrix or
  "linear" for linear cost.

Compute or check the structure of a cost matrix

An implementation of "Bundle Methods for Regularized Risk Minimization" (BMRM) inspired by work of Teo et al., JMLR 2010. This universal data mining framework is particularly useful for structured prediction, classification and regression on big data. The package implements L1 and L2 regularization for: linear SVM, multiclass SVM, f-beta optimization, ROC optimization, ordinal regression, quantile regression, epsilon insensitive regression, least mean square, logistic regression, least absolute deviation (see package examples). Other extensions are possible by implementing custom loss functions.

costMatrix: Compute or check the structure of a cost matrix

Description

Usage

Arguments

Value

See Also