Welcome to the OpAL Lab
The Optimization, Analytics, and Learning (OpAL) Lab is a research group at the Cheriton School of Computer Science. We work on machine learning for graphs and neural algorithmic reasoning. The former involves making predictions using multimodal datasets that combine features and relational information among entities. The latter focuses on using neural networks to solve problems that require algorithmic solutions. The two subjects often overlap.
We are located at the University of Waterloo, a vibrant technological hub with Concept, Velocity and Google around the block. We are also part of the Scientific Computation Group.
We are grateful for funding from the University of Waterloo and NSERC.
News
3 October 2024New paper: Positional Attention: Out-of-Distribution Generalization and Expressivity for Neural Algorithmic Reasoning. You can read a brief announcement of it here.
4 May 2024Our paper Analysis of Corrected Graph Convolutions was accepted at NeurIPS 2024.
3 September 2024US patent on Local Node Embeddings for Heterogeneous Graphs.
1 July 2024Aseem Baranwal will join the AI residency program at XTY Labs.
1 July 2024I wrote a blog at Medium on Random Data and Graph Neural Networks.