<< fda (X->Y_classes) Classification knnda >>

fact >> (X->Y_classes) Classification > forwda

forwda

forward discriminant analysis with cross-validation

Calling sequence

model = forwda(x,y_class,split,lv,(metric),(scale),(threshold))

Arguments

x:

a matrix (n x q) or a Div structure

y_class:

classification of the observations; a disjunctive matrix (n x nclass) or a conjunctive vector (n x 1) or Div structures

split:

for the cross-validation; a number of blocks, or a vector of dimensions (n x 1) identifying by a number the CV group of each observation

lv:

the number of latent variables or eigenvectors used to build the models

(scale):

the columns of x are always centered; scale defines the standardization

scale='c': no standardization

scale='cs': standardisation; the columns of x are divided by their standard deviation (by default)

(threshold):

the lowest value for attributing an observation to a class (by default= 1/nclass )

model.conf_cal_nobs:

the confusion matrix obtained with the calibration models,expressed as numbers of observations

model.conf_cal.d is an hyper-matrix of dimensions (nclass x nclass x lv)

model.conf_cal:

the confusion matrix obtained with the calibration models, expressed as percentages

model.conf_cal.d is an hyper-matrix of dimensions (nclass x nclass x lv)

model.conf_cv:

the confusion matrix obtained with the cross-validations, expressed as percentages

model.conf_cv.d is an hypermatrix of dimensions (nclass x nclass x lv)

model.err:

the percentage of classification errors for calibrations (1st column) and cross-validations (2nd column)

model.err.d is a matrix (lv x 2)

model.errbycl_cal:

the percentage of calibration error, for each class

model.errbycl_cal.d is a matrix (lv x nclass)

model.errbycl_cv:

the percentage of cross-validation error, for each class

model.pcent_errbycl_cv.d is a matrix (lv x nclass)

model.notclassed:

the percentage of not-classed observations (all predictions lower than the threshold)

model.notclassed.d is a vector (lv x 1)

model.notclassed_bycl:

the percentage of not-classed observations (all predictions lower than the threshold), for each class

model.notclassed_bycl.d is a matrix (lv x nclass)

model.method:

the discriminant method; here: 'forwda'

model.loadings:

the eigenvectors

model.loadings.d is a matrix of dimensions (q x lv)

model.classif_metric:

the metric used onto the scores for the calculation of distances between observations and groups

model.scale:

the applied standardisation

model.threshold:

the applied threshold

Examples

[result1]=forwda(x,y,30,20)
[result1]=forwda(x,y,30,20,1,'cs',0.01)

Authors


Report an issue
<< fda (X->Y_classes) Classification knnda >>