arXiv subjects

Liya Fu

Publications and source records attributed to Liya Fu.

6 recordsLinked to original sources

arXiv2026

Mixed Effects Mixture of Experts: Modeling Double Heterogeneous Trajectories

Linear mixed-effects model (LMM) is a cornerstone of longitudinal data analysis, but is limited to adeptly make heterogeneous analyses predictable under both group-specific fixed effects and subject-specific random effects. To address this challenge, we propose a novel statistical framework by using a large model prototype: a mixed effects mixture of experts model (MEMoE). This framework integrates the divide-and-conquer paradigm of Mixture of Experts Models with classical mixed-effect modeling. In the proposed MEMoE, each expert is a full LMM dedicated to capturing the longitudinal trajectory of a specific latent subpopulation, while another model gating function learns to route subjects to the most appropriate expert in a data-driven manner based on baseline covariates. We develop a robust inferential procedure for parameter estimation based on the Laplace Expectation-Maximization algorithm, with standard errors calibrated using robust sandwich estimators to account for potential model misspecification. Extensive simulation studies and an empirical application demonstrate that MEMoE outperforms both traditional single-population LMM and conventional Mixture of Experts models in terms of parameter recovery, classification accuracy, and overall model fit.

stat.ME↗

arXiv2025

Teleology-Driven Affective Computing: A Causal Framework for Sustained Well-Being

Affective computing has made significant strides in emotion recognition and generation, yet current approaches mainly focus on short-term pattern recognition and lack a comprehensive framework to guide affective agents toward long-term human well-being. To address this, we propose a teleology-driven affective computing framework that unifies major emotion theories (basic emotion, appraisal, and constructivist approaches) under the premise that affect is an adaptive, goal-directed process that facilitates survival and development. Our framework emphasizes aligning agent responses with both personal/individual and group/collective well-being over extended timescales. We advocate for creating a "dataverse" of personal affective events, capturing the interplay between beliefs, goals, actions, and outcomes through real-world experience sampling and immersive virtual reality. By leveraging causal modeling, this "dataverse" enables AI systems to infer individuals' unique affective concerns and provide tailored interventions for sustained well-being. Additionally, we introduce a meta-reinforcement learning paradigm to train agents in simulated environments, allowing them to adapt to evolving affective concerns and balance hierarchical goals - from immediate emotional needs to long-term self-actualization. This framework shifts the focus from statistical correlations to causal reasoning, enhancing agents' ability to predict and respond proactively to emotional challenges, and offers a foundation for developing personalized, ethically aligned affective systems that promote meaningful human-AI interactions and societal well-being.

cs.HC↗

arXiv2023

Optimal subsampling algorithm for the marginal model with large longitudinal data

Big data is ubiquitous in practices, and it has also led to heavy computation burden. To reduce the calculation cost and ensure the effectiveness of parameter estimators, an optimal subset sampling method is proposed to estimate the parameters in marginal models with massive longitudinal data. The optimal subsampling probabilities are derived, and the corresponding asymptotic properties are established to ensure the consistency and asymptotic normality of the estimator. Extensive simulation studies are carried out to evaluate the performance of the proposed method for continuous, binary and count data and with four different working correlation matrices. A depression data is used to illustrate the proposed method.

stat.ME↗

arXiv2021

Robust penalized empirical likelihood in high dimensional longitudinal data analysis

As an effective nonparametric method, empirical likelihood (EL) is appealing in combining estimating equations flexibly and adaptively for incorporating data information. To select important variables and estimating equations in the sparse high-dimensional model, we consider a penalized EL method based on robust estimating functions by applying two penalty functions for regularizing the regression parameters and the associated Lagrange multipliers simultaneously, which allows the dimensionalities of both regression parameters and estimating equations to grow exponentially with the sample size. A first inspection on the robustness of estimating equations contributing to the estimating equations selection and variable selection is discussed from both theoretical perspective and intuitive simulation results in this paper. The proposed method can improve the robustness and effectiveness when the data have underlying outliers or heavy tails in the response variables and/or covariates. The robustness of the estimator is measured via the bounded influence function, and the oracle properties are also established under some regularity conditions. Extensive simulation studies and a yeast cell data are used to evaluate the performance of the proposed method. The numerical results reveal that the robustness of sparse estimating equations selection fundamentally enhances variable selection accuracy when the data have heavy tails and/or include underlying outliers.

stat.ME↗

arXiv2020

Robust approach for variable selection with high dimensional Logitudinal data analysis

This paper proposes a new robust smooth-threshold estimating equation to select important variables and automatically estimate parameters for high dimensional longitudinal data. A novel working correlation matrix is proposed to capture correlations within the same subject. The proposed procedure works well when the number of covariates p increases as the number of subjects n increases. The proposed estimates are competitive with the estimates obtained with the true correlation structure, especially when the data are contaminated. Moreover, the proposed method is robust against outliers in the response variables and/or covariates. Furthermore, the oracle properties for robust smooth-threshold estimating equations under "large n, diverging p" are established under some regularity conditions. Extensive simulation studies and a yeast cell cycle data are used to evaluate the performance of the proposed method, and results show that our proposed method is competitive with existing robust variable selection procedures.

stat.ME↗

arXiv2020

An efficient Gehan-type estimation for the accelerated failure time model with clustered and censored data

In medical studies, the collected covariates usually contain underlying outliers. For clustered /longitudinal data with censored observations, the traditional Gehan-type estimator is robust to outliers existing in response but sensitive to outliers in the covariate domain, and it also ignores the within-cluster correlations. To take account of within-cluster correlations, varying cluster sizes, and outliers in covariates, we propose weighted Gehan-type estimating functions for parameter estimation in the accelerated failure time model for clustered data. We provide the asymptotic properties of the resulting estimators and carry out simulation studies to evaluate the performance of the proposed method under a variety of realistic settings. The simulation results demonstrate that the proposed method is robust to the outliers existing in the covariate domain and lead to much more efficient estimators when a strong within-cluster correlation exists. Finally, the proposed method is applied to a medical dataset and more reliable and convincing results are hence obtained.

stat.ME↗