Women in Computing events
The summer school is targeted at early career researchers, mostly PhD students. In addition, there will be two events for undergraduate and master students for which you can register here.
The events are as follows:
-
Getting-started tutorial on graph machine learning
-
Panel discussion on the topic of 'My journey into geometric machine learning', followed by a networking session
This event is hosted by Imperial College London Women in Computing Society.
Tutorial: Getting started on graph ML
Delivered by Dr Sarah Parisot
Sarah Parisot is a senior research scientist at Huawei Noah’s ark lab London. She is the first female team leader at Huawei London, leading the machine learning team which investigates topics ranging from deep generative models to neural architecture search. She has been involved in multiple projects involving to graph neural networks applications, notably proposing one of the first applications of graph CNNs to the medical imaging field.
Tutorial abstract:
This hands-on tutorial provides a step-by-step guide to implementing graph neural networks using PyTorch Geometric. We will discuss the key data structures and challenges inherent to graph structured data, from the construction and visualisation of graph structured objects, to the implementation of state of the art graph convolution methods as message passing tasks. Finally, we will build and train two simple models to solve standard graph and node classification problems.
Panel discussion: My journey into geometric ML
If you're a geometric ML career beginner, join us in this event "My journey into geometric ML". We have a panel of geometric ML practitioners who have gotten into the field from different paths. They will be sharing their personal experience and suggestions on how to get started in the field. If you've wondered whether a PhD is necessary, how the PhD journey is in this field, and how to get started, join us for an insightful discussion with our invited panellists!
Panellists
Linh Tran
Linh Tran is a Senior Research Scientist at the Autodesk AI lab and a final-year PhD student at Imperial College London, supervised by Prof. Maja Pantic. Her research interest focuses on probabilistic deep learning, in particular learning interpretable and human-controllable representations, quantifying uncertainty and robustness, and applying these ideas to unstructured data, 3D objects and images. During her PhD at Imperial College London, her research focused on learning disentangled representations for face analysis. Her work has been published in ICML, CVPR, ICCV and IJCV. She also spent time as an intern at Samsung Research and Google Brain, and has been awarded the Adobe Fellowship in 2019.
Michelle Li
Michelle is a PhD candidate in the Department of Biomedical Informatics at Harvard Medical School. Advised by Prof. Marinka Zitnik, Michelle is developing deep graph representation learning algorithms to characterize drugs at a single cell resolution, as well as to better leverage rich biomedical knowledge graphs for diagnosing rare diseases. She holds a B.S. in Mathematical and Computational Science from Stanford University, where she built bioinformatics tools to further interrogate the genetic mechanisms underlying antibiotic resistance. In her free time, Michelle enjoys playing the ukulele and teaching computer science to middle and high school students from underserved communities.
Kasia Janocha
Kasia is a Jagiellonian University Computer Science graduate. She currently works as a Machine Learning Engineer in Twitter’s Cortex Learning Methods team, which conducts theoretical and applied research to provide solutions for Twitter. She works on Graph Neural Networks as well as investigates applying unsupervised graph learning for detecting misinformation. Previously, she was enrolled in a PhD in Imperial College London, and was employed as a Software Engineer at Google.
Dr Meng Zhang
Meng Zhang is a postdoctoral researcher in Smart Geometry Processing Group at University College London, working with Prof. Niloy J. Mitra. Her current interests focus on deep learning and data-driven methods applied in the research fields of 3D modeling, rendering and animation, especially in hairs and garments which have complex dynamic structures with ever-changing fashion styles.