Big Data-month

Over the course of a month, Sticky will organise several activities in collaboration with ChipSoft. All of these activities are related to the same theme, being Big Data.

Talk from ChipSoft (Dutch)
by Mickel Reemer
Tuesday 29 May, 13:00, BBG 0.75
Mickel Reemer, Senior Technisch Consultant bij ChipSoft neemt je mee in de wereld van Big Data. Weten we eigenlijk wel wat het is en kunnen we er iets mee binnen de gezondheidszorg? Een vraagstuk dat volop in de actualiteit leeft en waar velen een mening over zullen hebben die niet onder stoelen of banken gestoken wordt. De overheid zet vol in op de ontwikkeling van eHealthoplossingen en kostenbesparing in de gezondheidszorg. Big Data kan hier een bijdrage aan leveren. Wat vind jij hiervan? Kom naar de lezing en laat je mening horen!

Visual Sensemaking of Big Data
by Stef van den Elzen
Thursday 31 May, 11:00, BBG 1.65

A powerful solution to go from data to insight is with the use of human-in-the-loop interactive visualizations. A different approach is to automatically build models with machine learning. However, both solutions are not sufficient to support sensemaking across disciplines and verticals. To fully support sensemaking we need an exploratory approach that enables users to freely ask questions combined with a system that keeps up with the speed of thinking. In this lecture, I will present the academic research and show industry examples that lead to our solution for visual sensemaking of big data!

Lunch will be available.

Deep Learning of Time for Video Data
by Efstratios Gavves
Wednesday 6 June, 11:00, BBG 1.65

In the modern years Deep Learning has been a great force of change for Science as well as the Industry. Most works have focused mainly on static and stationary data, such as images or video frames. Moving towards time-related domains, like video understanding, has been particularly challenging. In this talk, I will give a brief overview of Deep Learning on time-related domains, like videos, and present modern approaches on video activity and event recognition and visual object tracking. Further, I will present open new challenges in the direction of Temporal Deep Learning, towards non-stationary data.

Lunch will be available.

Learning from Evolving Data with Limited Supervision
by Georg Krempl
Thursday 7 June, 11:00, BBG 1.65

Machine learning has become widely used throughout commerce, science, and technology. However, the ever increasing volumes of data are contrasted by various constraints, such as limited supervision, processing or storage capacities. This requires techniques to optimise the use and allocation of these capacities. Such techniques include active and transfer learning. Active machine learning aims to provide techniques for selecting the most insightful information (e.g., label annotations of data instances) to be queried from oracles (e.g., domain experts supervising the learning process). Transfer machine learning techniques aim to compensate limited supervision in a target domain (e.g., a new classification task) by transferring knowledge from a better known source domain (e.g., a previously mastered classification task).

Lunch will be available.

Closing Drinks
Thursday 7 June, 17:00, Minnaert Mezzanine