singleIter {AdaSampling}

singleIter() applies a single iteraction of AdaSampling procedure. It returns the probabilities of all samples as being a positive (P) or negative (N) instance, as a two column data frame.

Description

Classification algorithms included are support vector machines (svm), k-nearest neighbours (knn), logistic regression (logit), linear discriminant analysis (lda), feature weighted knn (wKNN).

Usage

singleIter(Ps, Ns, dat, test = NULL, pos.probs = NULL,
  una.probs = NULL, classifier = "svm", sampleFactor, seed, weights)

Arguments

Ps

names (name as index) of positive examples

Ns

names (name as index) of negative examples

dat

training data matrix, without class labels.

test

test data matrix, without class labels. Training data matrix will be used for testing if this is NULL (default).

pos.probs

a numeric vector of containing probability of positive examples been positive

una.probs

a numeric vector of containing probability of negative or unannotated examples been negative

classifier

classification algorithm to be used for learning. Current options are support vector machine, "svm", k-nearest neighbour, "knn", logistic regression "logit", linear discriminant analysis "lda", and feature weighted knn, "wKNN".

sampleFactor

provides a control on the sample size for resampling.

seed

sets the seed.

weights

feature weights, required when using weighted knn.

References

Yang, P., Liu, W., Yang. J. (2017) Positive unlabeled learning via wrapper-based adaptive sampling. International Joint Conferences on Artificial Intelligence (IJCAI), 3272-3279

Yang, P., Ormerod, J., Liu, W., Ma, C., Zomaya, A., Yang, J.(2018) AdaSampling for positive-unlabeled and label noise learning with bioinformatics applications. IEEE Transactions on Cybernetics, doi:10.1109/TCYB.2018.2816984