Today Kasia and Kamil participated in the 1st Workshop on Online Learning from Uncertain Data Streams held in Padua, Italy. Kasia (our PI) was coorganizing it.
The topic of the workshop is closely related to research within our BIPOLAR project.
Applications in various domains (computer science, engineering, medicine, economy, etc.) are based on sensor data or depend on data transmission in the cloud.
Effective modeling approaches to address such a massive amount of dynamically-changing data in a feasible period of time was the main topic of this workshop.
Traditional modeling approaches for static datasets are very often insufficient or ineffective for online data streams due to the fact that fast recursive procedures are required to attend to narrow time
and memory constraints. Models must be updated (parametrically and structurally) to many types of changes of the data sources.
Moreover, data streams may carry statistical, possibilistic and fuzzy uncertainties that arise in specific technical and contextual domains, which need to be adequately addressed.
Finally, explainable models are needed in several domains in which the final users are non-technicians.
Thus, new methods to linguistically explain the reasoning behind the outcomes of a model are needed in order to trust and understand predictions.
Online Learning from Uncertain Data Streams (OLUD) workshop has addressed also the following open questions:
(i) how explainability can be handled in online learning?
(ii) how uncertainty can improve online learning?
(iii) how hybrid methods could be combined to exploit their benefits for online learning?
See more pictures here:
PHOTOS
Here is the webpage of the workshop: ( LINK )
BACK