triTraining {SSLR}

General Interface for Tri-training model

Description

Tri-training is a semi-supervised learning algorithm with a co-training style. This algorithm trains three classifiers with the same learning scheme from a reduced set of labeled examples. For each iteration, an unlabeled example is labeled for a classifier if the other two classifiers agree on the labeling proposed.

Usage

triTraining(learner)

Arguments

learner

model from parsnip package for training a supervised base classifier using a set of instances. This model need to have probability predictions (or optionally a distance matrix) and it's corresponding classes.

Details

Tri-training initiates the self-labeling process by training three models from the original labeled set, using the learner function specified. In each iteration, the algorithm detects unlabeled examples on which two classifiers agree with the classification and includes these instances in the enlarged set of the third classifier under certain conditions. The generation of the final hypothesis is produced via the majority voting. The iteration process ends when no changes occur in any model during a complete iteration.

Value

A list object of class "triTraining" containing:

model

The final three base classifiers trained using the enlarged labeled set.

model.index

List of three vectors of indexes related to the training instances used per each classifier. These indexes are relative to the y argument.

instances.index

The indexes of all training instances used to train the three models. These indexes include the initial labeled instances and the newly labeled instances. These indexes are relative to the y argument.

model.index.map

List of three vectors with the same information in model.index but the indexes are relative to instances.index vector.

classes

The levels of y factor.

pred

The function provided in the pred argument.

pred.pars

The list provided in the pred.pars argument.

x.inst

The value provided in the x.inst argument.

References

ZhiHua Zhou and Ming Li.
Tri-training: exploiting unlabeled data using three classifiers.
IEEE Transactions on Knowledge and Data Engineering, 17(11):1529-1541, Nov 2005. ISSN 1041-4347. doi: 10.1109/TKDE.2005. 186.

Examples

library(tidyverse)
library(tidymodels)
library(caret)
library(SSLR)

data(wine)

set.seed(1)
train.index <- createDataPartition(wine$Wine, p = .7, list = FALSE)
train <- wine[ train.index,]
test  <- wine[-train.index,]

cls <- which(colnames(wine) == "Wine")

#% LABELED
labeled.index <- createDataPartition(wine$Wine, p = .2, list = FALSE)
train[-labeled.index,cls] <- NA

#We need a model with probability predictions from parsnip
#https://tidymodels.github.io/parsnip/articles/articles/Models.html
#It should be with mode = classification

#For example, with Random Forest
rf <-  rand_forest(trees = 100, mode = "classification") %>%
  set_engine("randomForest")


m <- triTraining(learner = rf) %>% fit(Wine ~ ., data = train)


#Accuracy
predict(m,test) %>%
  bind_cols(test) %>%
  metrics(truth = "Wine", estimate = .pred_class)