Do not go gentle into that good night
Hi! I am a Founding member at Kumo AI. I received my Ph.D. and M.S. degrees from Department of Computer Science, Stanford University, advised by Prof. Jure Leskovec. I was supported by JPMC PhD Fellowship and Baidu Scholarship during my PhD. At Kumo AI, I aim to build a relational machine learning engine for cloud databases.
My research aims at developing data-driven methods to study our interconnected world. I investigate scientific and industrial problems through the lens of graph/relational data, and develop AI/ML solutions for these problems.
[Full image]
Here we introduce a novel relational multi-task learning setting where test data point may present auxiliary task labels. We develop MetaLink, where our key innovation is to build a knowledge graph that connects data points and tasks and thus allows us to leverage labels from auxiliary tasks.
[PDF] [Code]
[Full image]
Here we develop a class of message passing GNNs, named Identity-aware Graph Neural Networks (ID-GNNs), with greater expressive power than the 1-WL test.
ID-GNN offers a minimal but powerful solution to limitations of existing GNNs.
[PDF] [Code] [Webpage]
[Full image]
Here we define and systematically study the architectural design space for GNNs which consists of 315,000 different designs over 32 different predictive tasks.
We release GraphGym, a powerful platform for exploring different GNN designs and tasks.
[PDF] [Code] [Webpage]
[Full image]
Here, we propose GRAPE, a general framework for feature imputation and label prediction in the
presence of missing data. Our key innovation is to formulate the problem using a graph representation,
where observations and features are two types of nodes, and the observed feature values are attributed edges.
[PDF] [Code] [Webpage]
[Full image]
Here we systematically investigate how does the graph structure of neural networks affect their predictive performance.
Our work opens new directions for the design of neural architectures and the understanding on neural networks in general.
[PDF] [Code] [Video Recording] [Slides]
[Full image]
[Full image]
Here we propose DiffPool, a differentiable graph pooling module that can generate hierarchical representations of graphs and can be combined with various graph neural network architectures in an end-to-end fashion.
[PDF] [Code]
Here we propose Position-aware Graph Neural Networks (PGNNs), a new class of GNNs for computing
position-aware node embeddings which existing GNNs cannot represent.
[PDF] [Code] [Webpage] [Video Recording]
Here we propose Hierarchical Temporal Convolutional Networks (HierTCN), a hierarchical deep learning architecture that makes dynamic recommendations based on users' sequential multi-session interactions with items.
[PDF] [Code]
Here we propose GNNExplainer, the first general, model-agnostic approach for providing interpretable explanations for predictions of any GNN-based model on any graph-based
machine learning task.
[PDF] [Code] [Webpage]
Here we propose Hierarchically Aggregated computation Graphs (HAGs), a new GNN representation technique that explicitly avoids redundancy by managing intermediate aggre- gation results hierarchically and eliminates repeated computations and unnecessary data transfers in GNN training and inference.
[PDF]
Crop yield prediction is central in ensuring the food security. We introduce the first deep learning based method to predict crop yield purely based on publicly available remote sensing data.
[PDF] [Code] [Project Webpage]
This paper presents an effective simulation model for
multi-line metro systems based on the OD (origin-destination)
data and the network connection data.
[PDF]
