Computer Science > Machine Learning
[Submitted on 24 Nov 2020]
Title:Minimum Variance Embedded Auto-associative Kernel Extreme Learning Machine for One-class Classification
View PDFAbstract:One-class classification (OCC) needs samples from only a single class to train the classifier. Recently, an auto-associative kernel extreme learning machine was developed for the OCC task. This paper introduces a novel extension of this classifier by embedding minimum variance information within its architecture and is referred to as VAAKELM. The minimum variance embedding forces the network output weights to focus in regions of low variance and reduces the intra-class variance. This leads to a better separation of target samples and outliers, resulting in an improvement in the generalization performance of the classifier. The proposed classifier follows a reconstruction-based approach to OCC and minimizes the reconstruction error by using the kernel extreme learning machine as the base classifier. It uses the deviation in reconstruction error to identify the outliers. We perform experiments on 15 small-size and 10 medium-size one-class benchmark datasets to demonstrate the efficiency of the proposed classifier. We compare the results with 13 existing one-class classifiers by considering the mean F1 score as the comparison metric. The experimental results show that VAAKELM consistently performs better than the existing classifiers, making it a viable alternative for the OCC task.
References & Citations
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
Connected Papers (What is Connected Papers?)
CORE Recommender (What is CORE?)
IArxiv Recommender
(What is IArxiv?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.