NEWS
bst 0.3-23 (2020-11-09)
- editorial changes to Rd files
bst 0.3-22 (2020-10-19)
- change parallel computing codes in cross-validation
bst 0.3-19
- add detailed description on s value in man/bst_control.Rd
bst 0.3-18
- stop bst running if any fk=NULL and family="tgaussianDC", "thingeDC", "tbinomDC", "binomdDC", "texpoDC", "tpoissonDC", "thuberDC", "clossRMM", "clossMM", "glossMM", "qlossMM"))
bst 0.3-15 (2018-07-23)
- stop bst running if any fk=NULL and family="tgaussianDC", "thingeDC", "tbinomDC", "binomdDC", "texpoDC", "tpoissonDC", "thuberDC", "clossR", "clossRMM", "clossMM", "glossMM", "qlossMM"))
- importFrom("doParallel", "stopImplicitCluster")
- add stopImplicitCluster() after registerDoParallel
- change the offset value for family="huber", "thuber" or "lar". This is the initial value of boosting in bst function
- add a new parameter 'start' in bst_control, a logical value. If start=TRUE and fk in bst_control is a vector of values, then bst iterations begin with fk. Otherwise, bst iterations begin with the default values. This can be useful, for instance, in rbst for the MM boosting algorithm.
- drop bst-package.Rd in man directory
bst 0.3-14 (2016-09-21)
- add family="closs", "clossMM", "clossR", "clossRMM", "gloss", "glossMM", "qloss", "qlossMM" in bst function
- add rfamily="closs", "gloss", "qloss" in rbst function
- add family="closs", "clossMM" in mbst function
- add rfamily="closs" in rmbst function
- add vignette files for breast cancer data analysis
bst 0.3-13 (2016-02-28)
- recall terminating boosting if loss value inreases in bst function, but still terminating for rbst if loss value increases
- check if response is 1/-1 in bst and rbst for (margin based) binary classification problem
- modify cv.bst to be computationally efficient using predict function more properly
- add cv.rbst and cv.rmbst
- change type=c("risk", "misc") to type=c("loss", "error") in cross validation related functions, to be consistent with predict functions
- add cv.rbst
- add index.html to inst/doc
bst 0.3-12 (2016-01-04)
- add a data analysis example to khan.Rnw
- add a new vignette mcl.Rnw
- modify ex1data function
- fix a bug in predict.mbst when newdata=NULL and newy=NULL
bst 0.3-11 (2015-12-19)
- add parallel computing for cross-validation in cv.bst function
- add twintype argument for learner="ls" and twinboost=TRUE. If twintype=2, it does the same before and excluding version 0.3-10. If twintype=1, it does the same as version 0.3-10
- change argument "type" in bst and related functions to "threshold"
- add vignette file
- in function bst, boosting was terminated early if the fitted loss function value increased. This was changed and boosting only terminates for those loss functions related to truncated loss functions
- add inst/extdata and two vignettes
bst 0.3-10
- change the algorithm with respect to base class in mbst.R
- change bst function with twinboost=TRUE for learner="ls". In the previous versions, bst implemented the General Twin L2Boosting with general weak learner (Alg 1), rathter than Twin L2Boosting with componentwise linear least squares (Alg 2), cf: Buhlmann and Hothorn (2010). However, as Proposition 2 demonstrates, Alg 1 is not equal to Alg2 unless underlying regression model is orthonormal. The change is to match Alg 1.
bst 0.3-9
- add robust multi-class boosting rmbst.R
bst 0.3-8
- add family="expo" in bst function and rfamily="texpo" in rbst function for exponential loss and truncated exponential loss, respectively
- terminate boosting and robust boosting if loss value inreases
bst 0.3-7
- add intercept estimate for family=gaussian and center=FALSE
bst 0.3-6
- intercept estimated
- hingeloss and hingengra were modified, they are for family="hinge" since version 0.3-11. Before version 0.3-6, they are for family="hinge2".
- function rbst added: boosting plus difference of convex algorithm with truncated loss functions
bst 0.3-5
- for non-numeric responses in mada, they are now forced to be numeric
- fixed bugs in mada when xte and yte are missing
- in DESCRIPTION, change Imports: gbm to Depends: gbm, the results can be different for mada
bst 0.3-4 (2014-06-24)
- change output name fraction to mstop in cv.bst
bst 0.3-3 (2013-09-12)
- function name mhinge was changed to mhingeova (multi-class hinge one-vs-all)
- function name msvm was changed to mhingebst (multi-class hinge boosting)
bst 0.3-2 (2012-01-17)
- added trees size selection in multi-class HingeBoost msvm
bst 0.3-1 (2011-06-24)
- multi-class HingeBoost msvm
- add misclassification error as a criteria in cv.msvm for multiclass problem
bst 0.2 (2011-03-31)
- multi-class HingeBoost via one-vs-all method