School of Computer Science

Faculty of Science: High Dimensional Low Sample Sized Datasets (HDLSS)

22nd March 2017

Marcus Low
Postgraduate Administrator
Faculty of Science
University of Nottingham, Malaysian Campus


22 March 2017 (Venue: BA05)

Name : Sheena Leeza – School of Computer Science

Chairperson : Dr Iman Yi Liao

Evaluator : Dr Abdur Rakib

Title : High Dimensional Low Sample Sized Datasets (HDLSS)

Abstract :

High dimensional low sample size (HDLSS) datasets are commonly found in microarray data and medical imaging. With the increase in the number of data collected through improved communications among experts or data harvesting by agencies and businesses, there is a need to analyse and select data that is important to the task at hand. High dimensional small sample sized(HDLSS) datasets are datasets which contain many features but contain a limited number of samples. Some of the machine learning algorithms and tests may not work under the HDLSS framework as it was built mainly for when the number of samples are in abundance and exceeds the fixed dimensionality. This talk will discuss some of the properties of HDLSS datasets, problems which arise from it as well as methods that have been tested using HDLSS datasets.

School of Computer Science

The University of Nottingham Malaysia Campus
Jalan Broga, 43500 Semenyih
Selangor Darul Ehsan

telephone: +6 (03) 8924 8767
fax: +6 (03) 8924 8018

Make an enquiry