Events
Here you can find information about, and register to events organized or co-organized by NORA.
Feature selection is an important task in data analysis in general and especially with healthcare data where patients are often few, and features are many. Apart from prediction, interpretation of models is also of high importance and feature selection has a crucial role in this respect. By combining machine learning with component based multivariate statistical methods (chemometrics) one can design a workflow that helps medical staff gain deeper insights into their data.
In this webinar Oliver Tomic will show how RENT can help in finding stable biomarkers across subgroups of patients as well as giving a touch of personalised medicine by flagging patients that diverge from the rest.
Follow this link to register for the webinar:
Current methods for training AI systems, such as in deep learning, are not biologically plausible. They require tremendous amount of energy and data for training, and they lack explainability. By looking at how the brain works, and how biological neural networks communicate and process information, it may be possible to develop AI systems that are less artificial and more true to biology, and therefore closer to replicating some of the key features of animal brains.
In this presentation, Stefano Nichele will discuss ongoing research using in-vitro biological neural network recordings, and novel bio-inspired AI models and substrates.
Follow this link to register for the webinar:
