Introduction
This event has now been cancelled- we will let you know when we rearrange.
An event series for Turing-Roche partnership updates, knowledge sharing and new perspectives. Find out more about the series.
This event will explore concepts, method development and challenges of representing data in complex modern datasets.
About the event
In many applications, especially in biomedical and climate research, the volume and diversity of accessible data have reached unprecedented levels, offering a unique opportunity to gain deeper insights into these complex systems.
To characterise and address these challenges in modern datasets, various data representations have been developed, including low-dimensional projection, matrix factorisation, and graph representation. Among these, graph embeddings prove to be a remarkably powerful tool to provide insights into the geometry of the underlying data.
We will be hearing from Alessandro Barp, Senior Research Associate in the Turing-Roche Partnership who will be discussing the pitfalls and limitations of modern graph embedding algorithms, particularly in the context of missing data. He will also discuss some foundational concepts underpinning the Turing Theory and Method Challenge Fortnight event which will be run in the future.
