Author name cluster

Yijun Li

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

10 papers

1 author row

EAAI Journal 2026 Journal Article

A physics-guided neural network architecture for nonlinear hysteresis modeling of dielectric elastomer actuators

Hongfei Wang
Lei Ni
Linhai Huang
Yijun Li
Geng Wang

The input-output dynamics of dielectric elastomer actuators (DEAs) are inherently nonlinear and strongly hysteretic, posing significant challenges to accurate control unless rigorously modeled. To address this challenge, the paper proposes an innovative physics-guided neural network modeling framework. Specifically, a prior-physics knowledge-constrained driving mechanism is firstly designed by embedding the operational relationships of the Fractional-order Backlash-like differential equation as constraints on the neural network topology, achieving a clear one-to-one relationship between the equation's parameters and the neural network's weights. The resulting customized neural network model has a clear and transparent structure, marking the successful transformation of a Fractional-order Backlash-like differential equation model into a neural network representation for the first time. Then a gated recurrent unit (GRU) module is further integrated to compensate for unmodeled dynamic errors. The GRU's specially designed reset and update gates enable effective capture and processing of temporal dependencies. Experimental results demonstrate the proposed method's significant advantages in modeling the complex hysteresis behavior of DEAs: compared to traditional Fractional-order Backlash-like models, the average modeling error is reduced by approximately 30%, and peak-to-valley error is decreased to just 11% of its original value, indicating improved stability and accuracy. Furthermore, when compared to conventional GRU-based models, the proposed approach achieves a 30% reduction in runtime while maintaining comparable modeling precision, highlighting its superior overall performance and potential for practical deployment. This modeling strategy aims to provide a novel perspective for nonlinear hysteresis modeling of DEAs and other smart materials, thereby expanding the boundaries of physics-guided and deep learning integration.

Details DOI

AAAI Conference 2025 Conference Paper

Distributionally Robust Policy Evaluation and Learning for Continuous Treatment with Observational Data

Cheuk Hang Leung
Yiyan Huang
Yijun Li
Qi Wu

Using offline observational data for policy evaluation and learning allows decision-makers to evaluate and learn a policy that connects characteristics and interventions. Most existing literature has focused on either discrete treatment spaces or assumed no difference in the distributions between the policy-learning and policy-deployed environments. These restrict applications in many real-world scenarios where distribution shifts are present with continuous treatment. To overcome these challenges, this paper focuses on developing a distributionally robust policy under a continuous treatment setting. The proposed distributionally robust estimators are established using the Inverse Probability Weighting (IPW) method extended from the discrete one for policy evaluation and learning under continuous treatments. Specifically, we introduce a kernel function into the proposed IPW estimator to mitigate the exclusion of observations that can occur in the standard IPW method to continuous treatments. We then provide finite-sample analysis that guarantees the convergence of the proposed distributionally robust policy evaluation and learning estimators. The comprehensive experiments further verify the effectiveness of our approach when distribution shifts are present.

PDF Details DOI

FLAP Journal 2024 Journal Article

Algebraic Study of Substructural Fuzzy Epistemic Logics

Yongwei Yang
Yijun Li

This paper generalizes the notion of monadic residuated lattices to that of pseudo monadic residuated lattices. As monadic residuated lattices serve as algebraic models of modal logic S5(FLew ), we propose pseudo monadic resid- uated lattices as algebraic models of modal system KD45(FLew ). The main contributions of this paper are as follows: 1) we discuss the relationship be- tween pseudo monadic residuated lattices and other pseudo monadic algebraic structures, showing that it is a natural generalization of pseudo monadic BL- algebras, Bi-modal Gödel algebras and pseudo monadic algebras; 2) We provide a comprehensive characterization of pseudo monadic residuated lattices by con- sidering them as pairs of residuated lattices (L, B), where B represents a special case of a relatively complete subalgebra of L known as c-relatively complete. Furthermore, we establish a necessary and sufficient condition for a subalgebra to be c-relatively complete.

PDF Details

YNIMG Journal 2024 Journal Article

Effects of computerized working memory training on neuroplasticity in healthy individuals: A combined neuroimaging and neurotransmitter study

Peng Fang
Yuntao Gao
Yijun Li
Chenxi Li
Tian Zhang
Lin Wu
Yuanqiang Zhu
Yuanjun Xie

Working memory (WM) is an essential cognitive function that underpins various higher-order cognitive processes. Improving WM capacity through targeted training interventions has emergered as a potential approach for enhancing cognitive abilities. The present study employed an 8-week regimen of computerized WM training (WMT) to investigate its effect on neuroplasticity in healthy individuals, utilizing neuroimaging data gathered both before and after the training. The key metrics assessed included the amplitude of low-frequency fluctuations (ALFF), voxel-based morphometry (VBM), and the spatial distribution correlations of neurotransmitter. The results indicated that post-training, compared to baseline, there was a reduction in ALFF in the medial superior frontal gyrus and an elevation in ALFF in the left middle occipital gyrus within the training group. In comparison to the control group, the training group also exhibited decreased ALFF in the anterior cingulate cortex, angular gyrus, and superior parietal lobule, along with increased ALFF in the postcentral gyrus post-training. VBM analysis revealed a significant increase in gray matter volume (GMV) in the right dorsal superior frontal gyrus after the training period, compared to the initial baseline measurement. Furthermore, the training group showed GMV increases in the dorsal superior frontal gyrus, Rolandic operculum, precentral gyrus, and postcentral gyrus when compared to the control group. In addition, significant associations were identifed between neuroimaging measurements (AFLL and VBM) and the spatial patterns of neurotransmitters such as serotonin (5-HT), dopamine (DA), and N-methyl-D-aspartate (NMDA), providing insights into the underlying neurochemical processes. These findings clarify the neuroplastic changes caused by WMT, offering a deeper understanding of brain plasticity and highlighting the potential advantages of cognitive training interventions.

Details DOI

AAAI Conference 2024 Conference Paper

The Causal Impact of Credit Lines on Spending Distributions

Yijun Li
Cheuk Hang Leung
Xiangqian Sun
Chaoqun Wang
Yiyan Huang
Xing Yan
Qi Wu
Dongdong Wang

Consumer credit services offered by electronic commerce platforms provide customers with convenient loan access during shopping and have the potential to stimulate sales. To understand the causal impact of credit lines on spending, previous studies have employed causal estimators, (e.g., direct regression (DR), inverse propensity weighting (IPW), and double machine learning (DML)) to estimate the treatment effect. However, these estimators do not treat the spending of each individual as a distribution that can capture the range and pattern of amounts spent across different orders. By disregarding the outcome as a distribution, valuable insights embedded within the outcome distribution might be overlooked. This paper thus develops distribution valued estimators which extend from existing real valued DR, IPW, and DML estimators within Rubin’s causal framework. We establish their consistency and apply them to a real dataset from a large electronic commerce platform. Our findings reveal that credit lines generally have a positive impact on spending across all quantiles, but consumers would allocate more to luxuries (higher quantiles) than necessities (lower quantiles) as credit lines increase.

PDF Details DOI

NeurIPS Conference 2024 Conference Paper

Unveiling the Potential of Robustness in Selecting Conditional Average Treatment Effect Estimators

Yiyan Huang
Cheuk H. Leung
Siyi Wang
Yijun Li
Qi Wu

The growing demand for personalized decision-making has led to a surge of interest in estimating the Conditional Average Treatment Effect (CATE). Various types of CATE estimators have been developed with advancements in machine learning and causal inference. However, selecting the desirable CATE estimator through a conventional model validation procedure remains impractical due to the absence of counterfactual outcomes in observational data. Existing approaches for CATE estimator selection, such as plug-in and pseudo-outcome metrics, face two challenges. First, they must determine the metric form and the underlying machine learning models for fitting nuisance parameters (e. g. , outcome function, propensity function, and plug-in learner). Second, they lack a specific focus on selecting a robust CATE estimator. To address these challenges, this paper introduces a Distributionally Robust Metric (DRM) for CATE estimator selection. The proposed DRM is nuisance-free, eliminating the need to fit models for nuisance parameters, and it effectively prioritizes the selection of a distributionally robust CATE estimator. The experimental results validate the effectiveness of the DRM method in selecting CATE estimators that are robust to the distribution shift incurred by covariate shift and hidden confounders.

PDF Details DOI

IJCAI Conference 2023 Conference Paper

DeLELSTM: Decomposition-based Linear Explainable LSTM to Capture Instantaneous and Long-term Effects in Time Series

Chaoqun Wang
Yijun Li
Xiangqian Sun
Qi Wu
Dongdong Wang
Zhixiang Huang

Time series forecasting is prevalent in various real-world applications. Despite the promising results of deep learning models in time series forecasting, especially the Recurrent Neural Networks (RNNs), the explanations of time series models, which are critical in high-stakes applications, have received little attention. In this paper, we propose a Decomposition-based Linear Explainable LSTM (DeLELSTM) to improve the interpretability of LSTM. Conventionally, the interpretability of RNNs only concentrates on the variable importance and time importance. We additionally distinguish between the instantaneous influence of new coming data and the long-term effects of historical data. Specifically, DeLELSTM consists of two components, i. e. , standard LSTM and tensorized LSTM. The tensorized LSTM assigns each variable with a unique hidden state making up a matrix h(t), and the standard LSTM models all the variables with a shared hidden state H(t). By decomposing the H(t) into the linear combination of past information h(t-1) and the fresh information h(t)-h(t-1), we can get the instantaneous influence and the long-term effect of each feature. In addition, the advantage of linear regression also makes the explanation transparent and clear. We demonstrate the effectiveness and interpretability of DeLELSTM on three empirical datasets. Extensive experiments show that the proposed method achieves competitive performance against the baseline methods and provides a reliable explanation relative to domain knowledge.

PDF Details DOI

NeurIPS Conference 2020 Conference Paper

Few-shot Image Generation with Elastic Weight Consolidation

Yijun Li
Richard Zhang
Jingwan (Cynthia) Lu
Eli Shechtman

Few-shot image generation seeks to generate more data of a given domain, with only few available training examples. As it is unreasonable to expect to fully infer the distribution from just a few observations (e. g. , emojis), we seek to leverage a large, related source domain as pretraining (e. g. , human faces). Thus, we wish to preserve the diversity of the source domain, while adapting to the appearance of the target. We adapt a pretrained model, without introducing any additional parameters, to the few examples of the target domain. Crucially, we regularize the changes of the weights during this adaptation, in order to best preserve the information of the source dataset, while fitting the target. We demonstrate the effectiveness of our algorithm by generating high-quality results of different target domains, including those with extremely few examples (e. g. , 10). We also analyze the performance of our method with respect to some important factors, such as the number of examples and the similarity between the source and target domain.

PDF Details

IS Journal 2020 Journal Article

Stock Selection Model Based on Machine Learning with Wisdom of Experts and Crowds

Xianjiao Wu
Qiang Ye
Hong Hong
Yijun Li

Both stock recommendations from sell-side analysts and online user generated content from crowds have great significance in the stock market. We examine and compare different effects of analyst attitude and crowd sentiment on stock prices in this article with data from CSMAR. By estimating a multivariate linear regression model, we find that although the wisdom of both experts and crowds has impact on stock prices, the latter's impact on stock prices prevails. We also adopt LightGBM, a novel machine learning model, to predict stock trends based on empirical results. Portfolio returns of different models also suggest that crowd wisdom is more valuable for creating investment strategy than expert wisdom. And it is necessary to take the wisdom of both experts and crowds into consideration when making investment decision.

Details DOI

NeurIPS Conference 2017 Conference Paper

Universal Style Transfer via Feature Transforms

Yijun Li
Chen Fang
Jimei Yang
Zhaowen Wang
Xin Lu
Ming-Hsuan Yang

Universal style transfer aims to transfer arbitrary visual styles to content images. Existing feed-forward based methods, while enjoying the inference efficiency, are mainly limited by inability of generalizing to unseen styles or compromised visual quality. In this paper, we present a simple yet effective method that tackles these limitations without training on any pre-defined styles. The key ingredient of our method is a pair of feature transforms, whitening and coloring, that are embedded to an image reconstruction network. The whitening and coloring transforms reflect direct matching of feature covariance of the content image to a given style image, which shares similar spirits with the optimization of Gram matrix based cost in neural style transfer. We demonstrate the effectiveness of our algorithm by generating high-quality stylized images with comparisons to a number of recent methods. We also analyze our method by visualizing the whitened features and synthesizing textures by simple feature coloring.

PDF Details