Yujia Zheng (郑雨嘉)

I am a PhD student at CMU, fortunately advised by Prof. Kun Zhang. I am part of the CMU-CLeaR group. My research interests lie primarily in the linear span of causality and machine learning. Currently, I mainly study the following topics: causal discovery, causal representation learning, and structure learning. I am also interested in deep learning from a causal view.

Email  /  Google Scholar  /  LinkedIn

Causal-learn is an open-source Python package for causal discovery. It offers the implementations of up-to-date causal discovery methods as well as simple and intuitive APIs. Causal-learn is contributed by several groups and I am the lead coordinator. This project is under active development, and any comments or suggestions are welcome.

Documentation, GitHub, Paper

We present a general, unified local causal discovery method with linear non-Gaussian models, whether they are cyclic or acyclic.

We develop a Rank-based Latent Causal Discovery algorithm, RLCD, that can efficiently locate hidden variables, determine their cardinalities, and discover the entire causal structure over both measured and hidden ones.

We introduce causal-learn, an open-source Python library for causal discovery. We also briefly summarize different categories of methods in causal discovery and their use cases.

We establish a set of new identifiability results of nonlinear ICA in the general settings of undercompleteness, partial sparsity and source dependence, and flexible grouping structures.

We develop an identifiability theory that relies on second-order statistics without imposing further preconditions on the distribution of sources, by introducing novel assumptions on the connective structure from sources to observed variables.

We generalize the characterization of the conditional independence structure to handle general distributions for all data types, thus giving rise to a Markov network structure learning algorithm in one of the most general settings.

We prove the identifiability of nonlinear ICA with assumptions only on the mixing process, such as Structural Sparsity.

We show that under reasonable assumptions on the data generating process, as well as leveraging the principle of minimality, we can obtain partial identifiability of the changing and invariant parts of the generating process.

We introduce several strategies to improve the scalability of exact score-based methods in the linear Gaussian setting with theoretical guarantees under weaker assumptions.

We propose a new scenario named source free unsupervised graph domain adaptation. In this scenario, the only information we can leverage from the source domain is the well-trained source model, without any exposure to the source graph and its labels.

We introduce a new task, namely whole-page unbiased learning to rank, and propose a novel framework to automatically discover and mitigate the biases in an end-to-end manner.

We propose a brain network generation approach via modeling the connections among ROIs as DAGs to identify effective brain connectivities and predict the target in an end-to-end manner.

We propose an end-to-end framework based on task-aware brain connectivity DAG structure generation for fMRI analysis.

We present a novel lightweight recommendation paradigm that allows a well-trained recommender to be customized for arbitrary device-specific memory constraints without retraining.

We explore meta-learning to alleviate cold-start problem without any kind of side information.

We introduce Long-tail Recommendation into Session-based Recommendation.

We model the cross-session transitions via a channel-aware graph neural network.

We propose a multimedia dance video automatic scoring framework.

We propose a real-time fall detection method for the elderly.

We explore a general, completely nonparametric setting of causal representation learning from multiple distributions (arising from heterogeneous data or nonstationary time series), without assuming hard interventions behind distribution changes.

We investigate how various perturbations in the genome can affect the survival of patients diagnosed with breast cancer with causal principles.

We model user-item interactions as a heterogeneous graph which consists of various edge types.

We introduce Inter-session Item-level Interactions into Session-based Recommendation.

Publicity Chair of the 38th Conference on Uncertainty in Artificial Intelligence (UAI 2022).

Local Arrangements Chair of the 39th Conference on Uncertainty in Artificial Intelligence (UAI 2023).

International Conference on Learning Representations (ICLR), Tiny Papers Track.

Neural Information Processing Systems (NeurIPS).

International Conference on Machine Learning (ICML).

International Conference on Learning Representations (ICLR).

Uncertainty in Artificial Intelligence (UAI).

International Conference on Artificial Intelligence and Statistics (AISTATS).

Causal Learning and Reasoning (CLeaR).

International Joint Conference on Artificial Intelligence (IJCAI).

AAAI Conference on Artificial Intelligence (AAAI).

The IEEE / CVF Computer Vision and Pattern Recognition Conference (CVPR).

Empirical Methods in Natural Language Processing (EMNLP).

The ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD).

The ACM Web Conference (TheWebConf ).

The ACM Web Search and Data Mining Conference (WSDM).

European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML PKDD).

Learning on Graphs Conference (LoG).

Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD).

SIAM International Conference on Data Mining (SDM)

The International AAAI Conference on Web and Social Media (ICWSM).

Transactions on Machine Learning Research (TMLR).

IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI).

IEEE Transactions on Knowledge and Data Engineering (TKDE).

IEEE Transactions on Neural Networks and Learning Systems (TNNLS).

IEEE Transactions on Big Data (TBD).

IEEE Transactions on Information Systems (TOIS).

WSDM 2024 Best Paper Award (Honorable Mention).

NeurIPS 2023 Scholar Award.

Member of the final winning team in the NeurIPS 2022 CausalML Challenge. Ranked the 1st in the competition (the 1st, 1st, 1st, and 2nd, in the four tasks, respectively).

Member of the winning team in the Cancer Immunotherapy Grand Challenge. Ranked 9/972 in Challenge 2: proposing novel gene knockouts to maximize T cells' cancer-fighting capabilities.

NeurIPS 2022 Scholar Award.

NeurIPS 2022 Top Reviewer.

Summer Research Fellowship at EPFL (<2%).

Top 1% in multiple data competitions.

Tang Lixin Scholarship.

This awesome template is from here. Updated: 2023/7