Working with limited data from the perspective of research

Working with limited data from the perspective of research:
Signal classification and regression

When: Thursday, January 26, 18:00 - 20:00 CET
Where: Startit Center Belgrade, Savska 5

Our guest speaker at DATA SCIENCE CLUB in Belgrade will be Ivan Vajs, a junior research assistant at the Innovation Center of the School of Electrical Engineering in Belgrade and a Ph.D. student at the School of Electrical Engineering in Belgrade. We will hold our traditional networking session immediately after Ivan's talk!

Abstract

Machine learning algorithms, in combination with signal processing techniques, give outstanding results in many fields, especially when an abundance of data is available. However, working in niche fields often requires creativity and careful analysis in order to present adequate and interpretable results with often limited data. This happens very often in the fields of scientific research, where good quality and large datasets are hard to find.

The talk will present the experiences and challenges that come with using machine learning, focusing on signal classification and regression using limited datasets. We will cover two very different fields of research that are surprisingly similar from the perspective of data science: the detection of dyslexia using eye-tracking data and the calibration of low-cost air quality devices.

About our speaker

Ivan Vajs is a junior research assistant at the Innovation Center of the School of Electrical Engineering in Belgrade and a Ph.D. student at the School of Electrical Engineering in Belgrade. He has participated in several projects focusing on signal processing and Machine learning applications and has worked in the fields of biomedical engineering and air quality sensor calibration.