Hybrid Approach for Inductive Semi Supervised Learning using Label Propagation and Support Vector Machine

TL;DR

This paper introduces a hybrid semi-supervised learning method combining SVM and Label Propagation, demonstrating significant improvements over traditional methods across multiple datasets with faster training times when parallelized.

Contribution

The paper presents a novel hybrid approach that integrates SVM and Label Propagation for semi-supervised learning, with extensive experiments showing superior performance and reduced training time.

Findings

SVM outperforms Logistic Regression in semi-supervised tasks.

The hybrid approach nearly doubles the F-measure compared to Label Propagation.

Parallel implementation significantly reduces training time.

Abstract

Semi supervised learning methods have gained importance in today's world because of large expenses and time involved in labeling the unlabeled data by human experts. The proposed hybrid approach uses SVM and Label Propagation to label the unlabeled data. In the process, at each step SVM is trained to minimize the error and thus improve the prediction quality. Experiments are conducted by using SVM and logistic regression(Logreg). Results prove that SVM performs tremendously better than Logreg. The approach is tested using 12 datasets of different sizes ranging from the order of 1000s to the order of 10000s. Results show that the proposed approach outperforms Label Propagation by a large margin with F-measure of almost twice on average. The parallel version of the proposed approach is also designed and implemented, the analysis shows that the training time decreases significantly when parallel version is used.