arXiv subjects

Guosheng Yin

Publications and source records attributed to Guosheng Yin.

At least 19 recordsLinked to original sources

arXiv2026

Bridging Local and Population Causal Effects: A Proximal Instrumental Variable Approach

Instrumental variable (IV) methods address treatment endogeneity, but with non-compliance and heterogeneous treatment effects a binary instrument generally identifies the local average treatment effect (LATE) among compliers rather than the population average treatment effect (ATE). When treatment effects and compliance probabilities are heterogeneous and dependent through latent factors, the ATE need not be identified by IV variation alone. We develop a proximal instrumental variable framework that uses proxies for these factors to adjust for the compliance weighting in LATE and recover the ATE. We show that the ATE is identified through either an outcome bridge capturing latent treatment effect heterogeneity or a compliance bridge representing inverse compliance probabilities, and combining the two yields a doubly robust representation. We derive the efficient influence function under known and unknown instrument propensities. For estimation, we propose regularized kernel minimax bridge estimators, with orthogonalized conditional moments for estimated propensity, and a three-way sequential cross-fitting scheme. We establish projected bridge-risk bounds, and give conditions for asymptotic normality and semiparametric efficiency. The framework provides a proxy-based route from local IV effects to population causal effects under latent treatment effect and compliance heterogeneity.

stat.ME↗

arXiv2026

OmniCustom: Sync Audio-Video Customization Via Joint Audio-Video Generation Model

Existing mainstream video customization methods focus on generating identity-consistent videos based on given reference images and textual prompts. Benefiting from the rapid advancement of joint audio-video generation, this paper proposes a more compelling new task: sync audio-video customization, which aims to synchronously customize both video identity and audio timbre. Specifically, given a reference image $I^{r}$ and a reference audio $A^{r}$, this novel task requires generating videos that maintain the identity of the reference image while imitating the timbre of the reference audio, with spoken content freely specifiable through user-provided textual prompts. To this end, we propose OmniCustom, a powerful DiT-based audio-video customization framework that can synthesize a video following reference image identity, audio timbre, and text prompts all at once in a zero-shot manner. Our framework is built on three key contributions. First, identity and audio timbre control are achieved through separate reference identity and audio LoRA modules that operate through self-attention layers within the base audio-video generation model. Second, we introduce a contrastive learning objective alongside the standard flow matching objective. It uses predicted flows conditioned on reference inputs as positive examples and those without reference conditions as negative examples, thereby enhancing the model ability to preserve identity and timbre. Third, we train OmniCustom on our constructed large-scale, high-quality audio-visual human dataset. Extensive experiments demonstrate that OmniCustom outperforms existing methods in generating audio-video content with consistent identity and timbre fidelity. Project page: https://omnicustom-project.github.io/page/.

cs.SD↗

arXiv2026

One-sample survival tests in the presence of non-proportional hazards in oncology clinical trials

In oncology, conduct well-powered time-to-event randomized clinical trials may be challenging due to limited patietns number. Many designs for single-arm trials (SATs) have recently emerged as an alternative to overcome this issue. They rely on the (modified) one-sample log-rank test (OSLRT) under the proportional hazards to compare the survival curves of an experimental and an external control group. We extend Finkelstein's formulation of OSLRT as a score test by using a piecewise exponential model for early, middle and delayed treatment effects and an accelerated hazards model for crossing hazards. We adapt the restricted mean survival time based test and construct a combination test procedure (max-Combo) to SATs. The performance of the developed are evaluated through a simulation study. The score tests are as conservative as the OSLRT and have the highest power when the data generation matches the model underlying score tests. The max-Combo test is more powerful than the OSLRT whatever the scenarios and is thus an interesting approach as compared to a score test. Uncertainty on the survival curve estimated of the external control group and its model misspecification may have a significant impact on performance. For illustration, we apply the developed tests on real data examples.

stat.AP↗

arXiv2026

Delaunay Weighted Two-sample Test for High-dimensional Data by Incorporating Geometric Information

Two-sample hypothesis testing is a fundamental problem with various applications, which faces new challenges in the high-dimensional context. To mitigate the issue of the curse of dimensionality, high-dimensional data are typically assumed to lie on a low-dimensional manifold. To incorporate geometric information in the data, we propose to apply the Delaunay triangulation and develop the Delaunay weight to measure the geometric proximity among data points. In contrast to existing similarity measures that only utilize pairwise distances, the Delaunay weight can take both the distance and direction information into account. A detailed computation procedure is developed to learn the unknown manifold and approximate the Delaunay weight. We further propose a novel nonparametric test statistic using the Delaunay weight matrix. Asymptotic normality under the null and consistency under the alternative of the test statistic are developed. Applied on simulated data, the new test shows robustness to the learning of the unknown manifold and exhibits substantial power gain if the distributions differ directions. The proposed test also shows great power on a real dataset of mice protein expression levels.

stat.ME↗

arXiv2025

Cross-Modal Alignment via Variational Copula Modelling

Various data modalities are common in real-world applications (e.g., electronic health records, medical images and clinical notes in healthcare). It is essential to develop multimodal learning methods to aggregate various information from multiple modalities. The main challenge is how to appropriately align and fuse the representations of different modalities into a joint distribution. Existing methods mainly rely on concatenation or the Kronecker product, oversimplifying the interaction structure between modalities and indicating a need to model more complex interactions. Additionally, the joint distribution of latent representations with higher-order interactions is underexplored. Copula is a powerful statistical structure for modelling the interactions among variables, as it naturally bridges the joint distribution and marginal distributions of multiple variables. We propose a novel copula-driven multimodal learning framework, which focuses on learning the joint distribution of various modalities to capture the complex interactions among them. The key idea is to interpret the copula model as a tool to align the marginal distributions of the modalities efficiently. By assuming a Gaussian mixture distribution for each modality and a copula model on the joint distribution, our model can generate accurate representations for missing modalities. Extensive experiments on public MIMIC datasets demonstrate the superior performance of our model over other competitors. The code is available at https://github.com/HKU-MedAI/CMCM.

cs.LG↗

arXiv2025

Variational Polya Tree

Density estimation is essential for generative modeling, particularly with the rise of modern neural networks. While existing methods capture complex data distributions, they often lack interpretability and uncertainty quantification. Bayesian nonparametric methods, especially the \polya tree, offer a robust framework that addresses these issues by accurately capturing function behavior over small intervals. Traditional techniques like Markov chain Monte Carlo (MCMC) face high computational complexity and scalability limitations, hindering the use of Bayesian nonparametric methods in deep learning. To tackle this, we introduce the variational \polya tree (VPT) model, which employs stochastic variational inference to compute posterior distributions. This model provides a flexible, nonparametric Bayesian prior that captures latent densities and works well with stochastic gradient optimization. We also leverage the joint distribution likelihood for a more precise variational posterior approximation than traditional mean-field methods. We evaluate the model performance on both real data and images, and demonstrate its competitiveness with other state-of-the-art deep density estimation methods. We also explore its ability in enhancing interpretability and uncertainty quantification. Code is available at https://github.com/howardchanth/var-polya-tree.

cs.LG↗

arXiv2025

Amplifying Prominent Representations in Multimodal Learning via Variational Dirichlet Process

Developing effective multimodal fusion approaches has become increasingly essential in many real-world scenarios, such as health care and finance. The key challenge is how to preserve the feature expressiveness in each modality while learning cross-modal interactions. Previous approaches primarily focus on the cross-modal alignment, while over-emphasis on the alignment of marginal distributions of modalities may impose excess regularization and obstruct meaningful representations within each modality. The Dirichlet process (DP) mixture model is a powerful Bayesian non-parametric method that can amplify the most prominent features by its richer-gets-richer property, which allocates increasing weights to them. Inspired by this unique characteristic of DP, we propose a new DP-driven multimodal learning framework that automatically achieves an optimal balance between prominent intra-modal representation learning and cross-modal alignment. Specifically, we assume that each modality follows a mixture of multivariate Gaussian distributions and further adopt DP to calculate the mixture weights for all the components. This paradigm allows DP to dynamically allocate the contributions of features and select the most prominent ones, leveraging its richer-gets-richer property, thus facilitating multimodal feature fusion. Extensive experiments on several multimodal datasets demonstrate the superior performance of our model over other competitors. Ablation analysis further validates the effectiveness of DP in aligning modality distributions and its robustness to changes in key hyperparameters. Code is anonymously available at https://github.com/HKU-MedAI/DPMM.git

cs.LG↗

arXiv2025

Cluster-Level Sparse Multi-Instance Learning for Whole-Slide Images

Multi-Instance Learning (MIL) is pivotal for analyzing complex, weakly labeled datasets, such as whole-slide images (WSIs) in computational pathology, where bags comprise unordered collections of instances with sparse diagnostic relevance. Traditional MIL approaches, including early statistical methods and recent attention-based frameworks, struggle with instance redundancy and lack explicit mechanisms for discarding non-informative instances, limiting their robustness and interpretability. We propose Cluster-level Sparse MIL (csMIL), a novel framework that integrates global-local instance clustering, within-cluster attention, and cluster-level sparsity induction to address these challenges. Our csMIL first performs global clustering across all bags to establish $K$ cluster centers, followed by local clustering within each bag to assign cluster labels. Attention scores are computed within each cluster, and sparse regularization is applied to cluster weights, enabling the selective retention of diagnostically relevant clusters while discarding irrelevant ones. This approach enhances robustness to noisy instances, improves interpretability by identifying critical regions, and reduces computational complexity. Theoretical analysis demonstrates that csMIL requires $O(s log K)$ bags to recover $s$ relevant clusters, aligning with compressed sensing principles. Empirically, csMIL achieves state-of-the-art performance on two public histopathology benchmarks (CAMELYON16, TCGA-NSCLC).

cs.CV↗

arXiv2025

Effective sample size estimation based on concordance between p-value and posterior probability of the null hypothesis

Estimating the effective sample size (ESS) of a prior distribution is an age-old yet pivotal challenge, with great implications for clinical trials and various biomedical applications. Although numerous endeavors have been dedicated to this pursuit, most of them neglect the likelihood context in which the prior is embedded, thereby considering all priors as "beneficial". In the limited studies of addressing harmful priors, specifying a baseline prior remains an indispensable step. In this paper, by means of the elegant bridge between the p-value and the posterior probability of the null hypothesis, we propose a new ESS estimation method based on p-value in the framework of hypothesis testing, expanding the scope of existing ESS estimation methods in three key aspects: (i) We address the specific likelihood context of the prior, enabling the possibility of negative ESS values in case of prior-likelihood disconcordance; (ii) By leveraging the well-established bridge between the frequentist and Bayesian configurations under noninformative priors, there is no need to specify a baseline prior which incurs another criticism of subjectivity; (iii) By incorporating ESS into the hypothesis testing framework, our $p$-value ESS estimation method transcends the conventional one-ESS-one-prior paradigm and accommodates one-ESS-multiple-priors paradigm, where the sole ESS may reflect the collaborative impact of multiple priors in diverse contexts. Through comprehensive simulation analyses, we demonstrate the superior performance of the p-value ESS estimation method in comparison with existing approaches. Furthermore, by applying this approach to an expression quantitative trait loci (eQTL) data analysis, we show the effectiveness of informative priors in uncovering gene eQTL loci.

stat.ME↗

arXiv2025

Covariance Supervised Principal Component Analysis

Principal component analysis (PCA) is a widely used unsupervised dimensionality reduction technique in machine learning, applied across various fields such as bioinformatics, computer vision and finance. However, when the response variables are available, PCA does not guarantee that the derived principal components are informative to the response variables. Supervised PCA (SPCA) methods address this limitation by incorporating response variables into the learning process, typically through an objective function similar to PCA. Existing SPCA methods do not adequately address the challenge of deriving projections that are both interpretable and informative with respect to the response variable. The only existing approach attempting to overcome this, relies on a mathematically complicated manifold optimization scheme, sensitive to hyperparameter tuning. We propose covariance-supervised principal component analysis (CSPCA), a novel SPCA method that projects data into a lower-dimensional space by balancing (1) covariance between projections and responses and (2) explained variance, controlled via a regularization parameter. The projection matrix is derived through a closed-form solution in the form of a simple eigenvalue decomposition. To enhance computational efficiency for high-dimensional datasets, we extend CSPCA using the standard Nyström method. Simulations and real-world applications demonstrate that CSPCA achieves strong performance across numerous performance metrics.

stat.AP↗

arXiv2025

Feature Preserving Shrinkage on Bayesian Neural Networks via the R2D2 Prior

Bayesian neural networks (BNNs) treat neural network weights as random variables, which aim to provide posterior uncertainty estimates and avoid overfitting by performing inference on the posterior weights. However, the selection of appropriate prior distributions remains a challenging task, and BNNs may suffer from catastrophic inflated variance or poor predictive performance when poor choices are made for the priors. Existing BNN designs apply different priors to weights, while the behaviours of these priors make it difficult to sufficiently shrink noisy signals or they are prone to overshrinking important signals in the weights. To alleviate this problem, we propose a novel R2D2-Net, which imposes the R^2-induced Dirichlet Decomposition (R2D2) prior to the BNN weights. The R2D2-Net can effectively shrink irrelevant coefficients towards zero, while preventing key features from over-shrinkage. To approximate the posterior distribution of weights more accurately, we further propose a variational Gibbs inference algorithm that combines the Gibbs updating procedure and gradient-based optimization. This strategy enhances stability and consistency in estimation when the variational objective involving the shrinkage parameters is non-convex. We also analyze the evidence lower bound (ELBO) and the posterior concentration rates from a theoretical perspective. Experiments on both natural and medical image classification and uncertainty estimation tasks demonstrate satisfactory performance of our method.

cs.LG↗

arXiv2025

Causal Effect of Functional Treatment

We study the causal effect with a functional treatment variable, where practical applications often arise in neuroscience, biomedical sciences, etc. Previous research concerning the effect of a functional variable on an outcome is typically restricted to exploring correlation rather than causality. The generalized propensity score, which is often used to calibrate the selection bias, is not directly applicable to a functional treatment variable due to a lack of definition of probability density function for functional data. We propose three estimators for the average dose-response functional based on the functional linear model, namely, the functional stabilized weight estimator, the outcome regression estimator and the doubly robust estimator, each of which has its own merits. We study their theoretical properties, which are corroborated through extensive numerical experiments. A real data application on electroencephalography data and disease severity demonstrates the practical value of our methods.

stat.ME↗

arXiv2025

Bayesian analysis of restricted mean survival time adjusted for covariates using pseudo-observations

The difference in restricted mean survival time (RMST) is a clinically meaningful measure to quantify treatment effect in randomized controlled trials, especially when the proportional hazards assumption does not hold. Several frequentist methods exist to estimate RMST adjusted for covariates based on modeling and integrating the survival function. A more natural approach may be a regression model on RMST using pseudo-observations, which allows for a direct estimation without modeling the survival function. Only a few Bayesian methods exist, and each requires a model of the survival function. We developed a new Bayesian method that combines the use of pseudo-observations with the generalized method of moments. This offers RMST estimation adjusted for covariates without the need to model the survival function, making it more attractive than existing Bayesian methods. A simulation study was conducted with different time-dependent treatment effects (early, delayed, and crossing survival) and covariate effects, showing that our approach provides valid results, aligns with existing methods, and shows improved precision after covariate adjustment. For illustration, we applied our approach to a phase III trial in prostate cancer, providing estimates of the treatment effect on RMST, comparable to existing methods. In addition, our approach provided the effect of other covariates on RMST and determined the posterior probability of the difference in RMST exceeds any given time threshold for any covariate, allowing for nuanced and interpretable results.

stat.AP↗

arXiv2025

MVPortrait: Text-Guided Motion and Emotion Control for Multi-view Vivid Portrait Animation

Recent portrait animation methods have made significant strides in generating realistic lip synchronization. However, they often lack explicit control over head movements and facial expressions, and cannot produce videos from multiple viewpoints, resulting in less controllable and expressive animations. Moreover, text-guided portrait animation remains underexplored, despite its user-friendly nature. We present a novel two-stage text-guided framework, MVPortrait (Multi-view Vivid Portrait), to generate expressive multi-view portrait animations that faithfully capture the described motion and emotion. MVPortrait is the first to introduce FLAME as an intermediate representation, effectively embedding facial movements, expressions, and view transformations within its parameter space. In the first stage, we separately train the FLAME motion and emotion diffusion models based on text input. In the second stage, we train a multi-view video generation model conditioned on a reference portrait image and multi-view FLAME rendering sequences from the first stage. Experimental results exhibit that MVPortrait outperforms existing methods in terms of motion and emotion control, as well as view consistency. Furthermore, by leveraging FLAME as a bridge, MVPortrait becomes the first controllable portrait animation framework that is compatible with text, speech, and video as driving signals.

cs.CV↗

arXiv2025

Democratizing Large Language Model-Based Graph Data Augmentation via Latent Knowledge Graphs

Data augmentation is necessary for graph representation learning due to the scarcity and noise present in graph data. Most of the existing augmentation methods overlook the context information inherited from the dataset as they rely solely on the graph structure for augmentation. Despite the success of some large language model-based (LLM) graph learning methods, they are mostly white-box which require access to the weights or latent features from the open-access LLMs, making them difficult to be democratized for everyone as existing LLMs are mostly closed-source for commercial considerations. To overcome these limitations, we propose a black-box context-driven graph data augmentation approach, with the guidance of LLMs -- DemoGraph. Leveraging the text prompt as context-related information, we task the LLM with generating knowledge graphs (KGs), which allow us to capture the structural interactions from the text outputs. We then design a dynamic merging schema to stochastically integrate the LLM-generated KGs into the original graph during training. To control the sparsity of the augmented graph, we further devise a granularity-aware prompting strategy and an instruction fine-tuning module, which seamlessly generates text prompts according to different granularity levels of the dataset. Extensive experiments on various graph learning tasks validate the effectiveness of our method over existing graph data augmentation methods. Notably, our approach excels in scenarios involving electronic health records (EHRs), which validates its maximal utilization of contextual knowledge, leading to enhanced predictive performance and interpretability.

cs.LG↗

arXiv2024

Deep Nonparametric Inference for Conditional Hazard Function

We propose a novel deep learning approach to nonparametric statistical inference for the conditional hazard function of survival time with right-censored data. We use a deep neural network (DNN) to approximate the logarithm of a conditional hazard function given covariates and obtain a DNN likelihood-based estimator of the conditional hazard function. Such an estimation approach renders model flexibility and hence relaxes structural and functional assumptions on conditional hazard or survival functions. We establish the nonasymptotic error bound and functional asymptotic normality of the proposed estimator. Subsequently, we develop new one-sample tests for goodness-of-fit evaluation and two-sample tests for treatment comparison. Both simulation studies and real application analysis show superior performances of the proposed estimators and tests in comparison with existing methods.

stat.ME↗

arXiv2024

Diagnosis and Pathogenic Analysis of Autism Spectrum Disorder Using Fused Brain Connection Graph

We propose a model for diagnosing Autism spectrum disorder (ASD) using multimodal magnetic resonance imaging (MRI) data. Our approach integrates brain connectivity data from diffusion tensor imaging (DTI) and functional MRI (fMRI), employing graph neural networks (GNNs) for fused graph classification. To improve diagnostic accuracy, we introduce a loss function that maximizes inter-class and minimizes intra-class margins. We also analyze network node centrality, calculating degree, subgraph, and eigenvector centralities on a bimodal fused brain graph to identify pathological regions linked to ASD. Two non-parametric tests assess the statistical significance of these centralities between ASD patients and healthy controls. Our results reveal consistency between the tests, yet the identified regions differ significantly across centralities, suggesting distinct physiological interpretations. These findings enhance our understanding of ASD's neurobiological basis and offer new directions for clinical diagnosis.

q-bio.NC↗

arXiv2024

Multi-task Heterogeneous Graph Learning on Electronic Health Records

Learning electronic health records (EHRs) has received emerging attention because of its capability to facilitate accurate medical diagnosis. Since the EHRs contain enriched information specifying complex interactions between entities, modeling EHRs with graphs is shown to be effective in practice. The EHRs, however, present a great degree of heterogeneity, sparsity, and complexity, which hamper the performance of most of the models applied to them. Moreover, existing approaches modeling EHRs often focus on learning the representations for a single task, overlooking the multi-task nature of EHR analysis problems and resulting in limited generalizability across different tasks. In view of these limitations, we propose a novel framework for EHR modeling, namely MulT-EHR (Multi-Task EHR), which leverages a heterogeneous graph to mine the complex relations and model the heterogeneity in the EHRs. To mitigate the large degree of noise, we introduce a denoising module based on the causal inference framework to adjust for severe confounding effects and reduce noise in the EHR data. Additionally, since our model adopts a single graph neural network for simultaneous multi-task prediction, we design a multi-task learning module to leverage the inter-task knowledge to regularize the training process. Extensive empirical studies on MIMIC-III and MIMIC-IV datasets validate that the proposed method consistently outperforms the state-of-the-art designs in four popular EHR analysis tasks -- drug recommendation, and predictions of the length of stay, mortality, and readmission. Thorough ablation studies demonstrate the robustness of our method upon variations to key components and hyperparameters.

cs.LG↗