Self-Trained LMT for Semisupervised Learning

Abstract

The most important asset of semisupervised classification methods is the use of available unlabeled data combined with a clearly smaller set of labeled examples, so as to increase the classification accuracy compared with the default procedure of supervised methods, which on the other hand use only the labeled data during the training phase. Both the absence of automated mechanisms that produce labeled data and the high cost of needed human effort for completing the procedure of labelization in several scientific domains rise the need for semisupervised methods which counterbalance this phenomenon. In this work, a self-trained Logistic Model Trees (LMT) algorithm is presented, which combines the characteristics of Logistic Trees under the scenario of poor available labeled data. We performed an in depth comparison with other well-known semisupervised classification methods on standard benchmark datasets and we finally reached to the point that the presented technique had better accuracy in most cases.

Semi-supervised Segmentation of Salt Bodies in Seismic Images Using an Ensemble of Convolutional Neural Networks

“…After that, the classifier is retrained with its own most confident predictions, together with initially provided labeled examples. However existing self-training approaches [12,51,32,31] are based on hand-crafted features which are much more limited than the features learned by CNNs. [27] and [44] use CNNs in the self-training framework, but they apply it to relatively simple classification datasets like MNIST [26] and CIFAR-10 [24].…”

Distant Supervision with Transductive Learning for Adverse Drug Reaction Identification from Electronic Medical Records

“…As the first parameter, recent works [ 55 – 57 ] have proposed various predictive models, such as generative models [ 22 , 58 ], low-density separation models [ 59 ], and graph-based models [ 60 ]. For the second parameter, at least two alternatives, namely self-training [ 61 , 62 ] or cotraining [ 63 ], can be applied to assign a label to an unlabeled instance by either one single predictive model or multiple ensemble predictive models. The last parameter concerns with how to handle test instances, where two choices are (i) to manipulate the test instances separately from the unlabeled instances (inductive learning) or (ii) to treat them as unlabeled instances in the training step (transductive learning).…”

Review on Predicting Students’ Graduation Time Using Machine Learning Algorithms

“…There are several categories that fall under semi-supervised learning algorithms such as Self-Training, Generative Model and Transudative Support Vector Machine. Self-Training is a wrapper algorithm where the labeled data are trained by the classifier and later will classify unlabelled data [38]. The unlabelled data that is associated with the highest confidence score will be added to the training set.…”

“…The unlabelled data that is associated with the highest confidence score will be added to the training set. As stated in [38], self-Training is known as the simplest algorithm in semi-supervised learning but still gives a good solution to the classification problem. Contrary with Self-Training, Generative Model has difficulties in providing good solutions to the classification problem, especially when the unlabelled data is more than labeled data, as it has a problem to balance the effect of unlabelled and labeled data.…”