1 code implementation • 6 May 2024 • Weilin Chen, Ruichu Cai, Zeqin Yang, Jie Qiao, Yuguang Yan, Zijian Li, Zhifeng Hao
Based on the condition, we devise an end-to-end causal effect estimator by transforming the identified theoretical condition into a targeted loss.
no code implementations • 24 Feb 2024 • Zijian Li, Ruichu Cai, Haiqin Huang, Sili Zhang, Yuguang Yan, Zhifeng Hao, Zhenghua Dong
Existing model-based interactive recommendation systems are trained by querying a world model to capture the user preference, but learning the world model from historical logged data will easily suffer from bias issues such as popularity bias and sampling bias.
no code implementations • 20 Feb 2024 • Zijian Li, Ruichu Cai, Zhenhui Yang, Haiqin Huang, Guangyi Chen, Yifan Shen, Zhengming Chen, Xiangchen Song, Zhifeng Hao, Kun Zhang
To solve this problem, we propose to learn IDentifiable latEnt stAtes (IDEA) to detect when the distribution shifts occur.
no code implementations • 17 Feb 2024 • Xiaolu Wang, Zijian Li, Shi Jin, Jun Zhang
Federated learning (FL) is an emerging distributed training paradigm that aims to learn a common global model without exchanging or transferring the data that are stored locally at different clients.
no code implementations • 14 Feb 2024 • Xuexin Chen, Ruichu Cai, Kaitao Zheng, Zhifan Jiang, Zhengting Huang, Zhifeng Hao, Zijian Li
Under mild conditions, we show that the invariant subgraph can be extracted by minimizing an upper bound, which is built on the theoretical advance of probability of necessity and sufficiency.
no code implementations • 13 Feb 2024 • Xuexin Chen, Ruichu Cai, Zhengting Huang, Yuxuan Zhu, Julien Horwood, Zhifeng Hao, Zijian Li, Jose Miguel Hernandez-Lobato
We investigate the problem of explainability in machine learning. To address this problem, Feature Attribution Methods (FAMs) measure the contribution of each feature through a perturbation test, where the difference in prediction is compared under different perturbations. However, such perturbation tests may not accurately distinguish the contributions of different features, when their change in prediction is the same after perturbation. In order to enhance the ability of FAMs to distinguish different features' contributions in this challenging setting, we propose to utilize the probability (PNS) that perturbing a feature is a necessary and sufficient cause for the prediction to change as a measure of feature importance. Our approach, Feature Attribution with Necessity and Sufficiency (FANS), computes the PNS via a perturbation test involving two stages (factual and interventional). In practice, to generate counterfactual samples, we use a resampling-based approach on the observed samples to approximate the required conditional distribution. Finally, we combine FANS and gradient-based optimization to extract the subset with the largest PNS. We demonstrate that FANS outperforms existing feature attribution methods on six benchmarks.
no code implementations • 20 Dec 2023 • Zijian Li, Zhihui Wang
Generative Adversarial Networks (GANs) have become a ubiquitous technology for data generation, with their prowess in image generation being well-established.
1 code implementation • 8 Nov 2023 • Zijian Li, Zunhong Xu, Ruichu Cai, Zhenhui Yang, Yuguang Yan, Zhifeng Hao, Guangyi Chen, Kun Zhang
Specifically, we first formulate the data generation process from the atom level to the molecular level, where the latent space is split into SI substructures, SR substructures, and SR atom variables.
1 code implementation • NeurIPS 2023 • Zijian Li, Ruichu Cai, Guangyi Chen, Boyang Sun, Zhifeng Hao, Kun Zhang
To mitigate the need for these strict assumptions, we propose a subspace identification theory that guarantees the disentanglement of domain-invariant and domain-specific variables under less restrictive constraints regarding domain numbers and transformation properties, thereby facilitating domain adaptation by minimizing the impact of domain shifts on invariant variables.
1 code implementation • 8 Sep 2023 • Haochun Wang, Sendong Zhao, Zewen Qiang, Zijian Li, Nuwa Xi, Yanrui Du, MuZhen Cai, Haoqiang Guo, Yuhan Chen, Haoming Xu, Bing Qin, Ting Liu
To address this challenge, we propose knowledge-tuning, which leverages structured medical knowledge bases for the LLMs to grasp domain knowledge efficiently and facilitate reliable response generation.
no code implementations • 30 Aug 2023 • Zijian Li, Zehong Lin, Jiawei Shao, Yuyi Mao, Jun Zhang
However, devices often have non-independent and identically distributed (non-IID) data, meaning their local data distributions can vary significantly.
no code implementations • 9 Aug 2023 • Zijian Li, Yuchang Sun, Jiawei Shao, Yuyi Mao, Jessie Hui Wang, Jun Zhang
For better privacy preservation, we propose a hard feature augmentation method to transfer real features towards the decision boundary, with which the synthetic data not only improve the model generalization but also erase the information of real features.
no code implementations • 20 Jul 2023 • Jiawei Shao, Zijian Li, Wenqiang Sun, Tailin Zhou, Yuchang Sun, Lumin Liu, Zehong Lin, Yuyi Mao, Jun Zhang
Without data centralization, FL allows clients to share local information in a privacy-preserving manner.
no code implementations • 25 Jun 2023 • Yuequn Liu, Ruichu Cai, Wei Chen, Jie Qiao, Yuguang Yan, Zijian Li, Keli Zhang, Zhifeng Hao
assumption is often violated due to the inherent dependencies among the event sequences.
1 code implementation • 12 Jul 2022 • Xubin Zhong, Changxing Ding, Zijian Li, Shaoli Huang
Specifically, we shift the GT bounding boxes of each labeled human-object pair so that the shifted boxes cover only a certain portion of the GT ones.
no code implementations • 11 Jun 2022 • Zijian Li, Jiawei Shao, Yuyi Mao, Jessie Hui Wang, Jun Zhang
A combination of the local private dataset and synthetic dataset with confident pseudo labels leads to nearly identical data distributions among clients, which improves the consistency among local models and benefits the global aggregation.
no code implementations • 7 May 2022 • Zijian Li, Ruichu Cai, Jiawei Chen, Yuguan Yan, Wei Chen, Keli Zhang, Junjian Ye
Based on this inspiration, we investigate the domain-invariant unweighted sparse associative structures and the domain-variant strengths of the structures.
1 code implementation • CVPR 2022 • Xin Lin, Changxing Ding, Yibing Zhan, Zijian Li, DaCheng Tao
Despite their effectiveness, however, current SGG methods only assume scene graph homophily while ignoring heterophily.
1 code implementation • 13 Jan 2022 • Ruichu Cai, Fengzhu Wu, Zijian Li, Jie Qiao, Wei Chen, Yuexing Hao, Hao Gu
By explicitly Reconstructing Exposure STrategies (REST in short), we formalize the recommendation problem as the counterfactual reasoning and propose the debiased social recommendation method.
1 code implementation • 30 Dec 2021 • Xuexin Chen, Ruichu Cai, Yuan Fang, Min Wu, Zijian Li, Zhifeng Hao
However, standard GNNs in the neighborhood aggregation paradigm suffer from limited discriminative power in distinguishing \emph{high-order} graph structures as opposed to \emph{low-order} structures.
1 code implementation • NeurIPS 2021 • Petar Stojanov, Zijian Li, Mingming Gong, Ruichu Cai, Jaime Carbonell, Kun Zhang
We provide reasoning why when the supports of the source and target data from overlap, any map of $X$ that is fixed across domains may not be suitable for domain adaptation via invariant features.
1 code implementation • 14 Nov 2021 • Zijian Li, Ruichu Cai, Fengzhu Wu, Sili Zhang, Hao Gu, Yuexing Hao, Yuguang
To achieve this, we firstly formalize sequential recommendation as a problem to estimate conditional probability given temporal dynamic heterogeneous graphs and user behavior sequences.
1 code implementation • 5 Nov 2021 • Zijian Li, Ruichu Cai, Tom Z. J Fu, Zhifeng Hao, Kun Zhang
In order to address these challenges, we analyze variational conditional dependencies in time-series data and find that the causal structures are usually stable among domains, and further raise the causal conditional shift assumption.
no code implementations • 14 Jun 2021 • Ruichu Cai, Fengzhu Wu, Zijian Li, Pengfei Wei, Lingling Yi, Kun Zhang
Based on this assumption, we propose a disentanglement-based unsupervised domain adaptation method for the graph-structured data, which applies variational graph auto-encoders to recover these latent variables and disentangles them via three supervised learning modules.
1 code implementation • 9 May 2021 • Changjian Shui, Zijian Li, Jiaqi Li, Christian Gagné, Charles Ling, Boyu Wang
Multi-source domain adaptation aims at leveraging the knowledge from multiple tasks for predicting a related target domain.
no code implementations • 1 Jan 2021 • Changjian Shui, Zijian Li, Jiaqi Li, Christian Gagné, Charles Ling, Boyu Wang
We study the label shift problem in multi-source transfer learning and derive new generic principles to control the target generalization risk.
1 code implementation • 22 Dec 2020 • Zhifeng Hao, Di Lv, Zijian Li, Ruichu Cai, Wen Wen, Boyan Xu
In the proposed framework, the domain-specific information is integrated with the domain-specific latent variables by using a domain predictor.
no code implementations • 22 Dec 2020 • Ruichu Cai, Jiawei Chen, Zijian Li, Wei Chen, Keli Zhang, Junjian Ye, Zhuozhang Li, Xiaoyan Yang, Zhenjie Zhang
To reduce the difficulty in the discovery of causal structure, we relax it to the sparse associative structure and propose a novel sparse associative structure alignment model for domain adaptation.
1 code implementation • 22 Dec 2020 • Ruichu Cai, Zijian Li, Pengfei Wei, Jie Qiao, Kun Zhang, Zhifeng Hao
Different from previous efforts on the entangled feature space, we aim to extract the domain invariant semantic information in the latent disentangled semantic representation (DSR) of the data.
no code implementations • ACL 2020 • Ruichu Cai, Zhihao Liang, Boyan Xu, Zijian Li, Yuexing Hao, Yao Chen
Existing leading code comment generation approaches with the structure-to-sequence framework ignores the type information of the interpretation of the code, e. g., operator, string, etc.
no code implementations • 30 Nov 2019 • Jie Qiao, Zijian Li, Boyan Xu, Ruichu Cai, Kun Zhang
The challenge of learning disentangled representation has recently attracted much attention and boils down to a competition using a new real world disentanglement dataset (Gondal et al., 2019).
no code implementations • 13 Oct 2019 • Zijian Li, Ruichu Cai, Kok Soon Chai, Hong Wei Ng, Hoang Dung Vu, Marianne Winslett, Tom Z. J. Fu, Boyan Xu, Xiaoyan Yang, Zhenjie Zhang
However, the mainstream domain adaptation methods cannot achieve ideal performance on time series data, because most of them focus on static samples and even the existing time series domain adaptation methods ignore the properties of time series data, such as temporal causal mechanism.
no code implementations • 16 Nov 2017 • Ruichu Cai, Boyan Xu, Xiaoyan Yang, Zhenjie Zhang, Zijian Li, Zhihao Liang
These techniques help the neural network better focus on understanding semantics of operations in natural language and save the efforts on SQL grammar learning.