SAFE-SEMI SUPERVISED FRAME WORK THROUGH MACHINE LEARNING

---

S.Jagadeesh
Professor, Department of Electronics and Communication Engineering, Sridevi Women’s Engineering College, Hyderabad, India, jaga.ssjec@gmail.com

Aerva Bhargavi
U.G. Student, Department of Electronics and Communication Engineering, Sridevi Women’s Engineering College, Hyderabad, India

Ruchitha Reddy Kuntla
U.G. Student, Department of Electronics and Communication Engineering, Sridevi Women’s Engineering College, Hyderabad, India

Farheen Begum
U.G. Student, Department of Electronics and Communication Engineering, Sridevi Women’s Engineering College, Hyderabad, India

Vol 17, No 07 ( 2022 )   |  DOI: 10.5281/zenodo.6877693   |   Licensing: CC 4.0   |   Pg no: 166-177   |   Published on: 25-07-2022

---

Abstract
Semi-supervised learning (SSL) is a machine learning algorithm totally depends on manifold regularization technique. We have different approaches in manifold regularization but we mostly use graphical approach. Though, the execution of Semi-supervised learning (SSL) depends on the construction of manifold graph and the safety degrees of unlabeled samples. As the SSL machine learning totally depends on the manifold graph (MR) and degrees of unlabeled samples i.e, those are usually constructed before the classification and fixed during the process of classification learning, then the results are independent with the classification. Focusing on the above problems, we proposed a unified algorithm called adaptive safe semi-supervised learning frame work. Here in this method construction of manifold graph and calculates the safety degrees of unlabeled samples. We will optimize the weights of manifold graph and degree of unlabeled samples in learning process and then it calculates in advance. At last, we develop and implement a adaptive safe semi-supervised extreme machine learning. Here the performance of this algorithm is effective, reliable and we get more accuracy as compared to others.

Keywords:
Semi-supervised learning (SSL), extreme learning machine, adaptive safety degree, adaptive graph, manifold regularization (MR).

---

Download Full Article PDF