Zihan Liao

Zihan Liao

Ph.D., Zhejiang University. Reliability estimation, Bayesian updating, machine learning and spectral analysis.

Zihan Liao

Reliability estimation · Bayesian updating · Machine Learning · Spectral analysis

My research focuses on developing general-purpose Monte Carlo simulation (MCS) algorithms for reliability estimation, Bayesian updating, and optimization. I am also interested in applying machine learning to engineering data and chemical spectra.

The research and data collection underlying my publications were carried out under the guidance of Professor Binbin Li at Zhejiang University. I gratefully acknowledge his support and guidance.

Contact: Please use 11912093@zju.edu.cn for all correspondence. My former email address, zihanliao@intl.zju.edu.cn, is no longer in use.

Education

  • 2025
    Ph.D., Zhejiang University
  • 2019
    Bachelor's degree, Nanchang University

Publications

Journal Articles

  1. 2026

    Tailored Deep Learning–Assisted In Situ SERS: Overcoming Surface Irregularities-Induced Large Signal Variation on Biological Tissues

    Ling Guo, Zihan Liao, Tianxi Yang*

    Analytical Chemistry, 98(25), 18645–18659

    Paper Code (MATLAB) Code (Python)

  2. 2026

    Sequential multiple importance sampling for robust and efficient (possibly high-dimensional) reliability estimation

    Weili Xia, Zihan Liao*, Binbin Li

    Mechanical Systems and Signal Processing, 259, 114798

    Paper Code (MATLAB) Code (Python)

  3. 2026

    Sequential multiple importance sampling for multi-modal Bayesian inference

    Binbin Li, Xiao He, Zihan Liao*

    Mechanical Systems and Signal Processing, 245, 113788

    Paper arXiv Code (MATLAB) Code (Python)

  4. 2025

    From likelihood to limit state: A reliability-inspired framework for Bayesian evidence estimation and high-dimensional sampling

    Zihan Liao, Binbin Li, Hua-Ping Wan*

    Mechanical Systems and Signal Processing, 237, 112969

    Paper arXiv Code (MATLAB) Code (Python)

  5. 2025

    Relaxed subset simulation for reliability estimation

    Binbin Li, Weili Xia, Zihan Liao*

    Reliability Engineering & System Safety, 264, 111302

    Paper Code (MATLAB) Code (Python)

  6. 2025

    Enhanced generalized subset simulation with multiple importance sampling for reliability estimation

    Weili Xia, Zihan Liao*

    Computers & Structures, 313, 107741

    Paper Code (MATLAB) Code (Python)

  7. 2025

    Bayesian mixture of factor analyzers for structural damage detection under varying environmental conditions

    Binbin Li, Yulong Zhang, Zihan Liao, Zhilin Xue*

    Journal of Infrastructure Intelligence and Resilience, 4(2), 100140

    Paper

  8. 2024

    Enhancing subset simulation through Bayesian inference

    Zihan Liao*, Xiao He, Weili Xia

    Computer Methods in Applied Mechanics and Engineering, 432, 117392

    Paper

  9. 2024

    An investigation into Markov chain Monte Carlo algorithms for Subset simulation: Emphasizing uncertainty analysis

    Zihan Liao*, Weili Xia, Xiao He

    Computers & Structures, 294, 107268

    Paper

  10. 2023

    子集模拟的混合采样算法

    廖子涵, 李宾宾*

    计算力学学报 (Chinese Journal of Computational Mechanics), 40(5), 693–700

  11. 2022

    拉索模态参数全过程自动化识别方法

    安永辉, 薛志林, 廖子涵, 李宾宾

    土木工程学报 (China Civil Engineering Journal), 56(7), 82–91

Conference Papers

  1. 2025

    Directional subset simulation for reliability estimation

    Weili Xia, Binbin Li, Zihan Liao

    14th International Conference on Structural Safety and Reliability (ICOSSAR)

  2. 2022

    Iterated subset simulation for Bayesian analysis

    Zihan Liao, Li Binbin*

    ICOSSAR 2021–2022, Shanghai

  3. 2022

    A strategy for optimal sensor placement in modal signal-to-noise ratio-based structural health monitoring

    Dahai Xiao, Zihan Liao, Binbin Li*

    2022 8th International Conference on Hydraulic and Civil Engineering (ICHCE), Xian, 1–8

    Paper

  4. 2020

    Adaptive stratified sampling using subset simulation for Bayesian analysis

    Zihan Liao, Li Binbin*

    7th International Symposium on Reliability Engineering and Risk Management (ISRERM), Beijing

Preprints

  1. 2026

    Understanding the geometry of high dimensional Bayesian inference via projected one-dimensional measures

    Binbin Li, Mingzhu Chen, Sin-Chi Kuok, Zihan Liao, Armen Der Kiureghian

    SSRN preprint

    Paper

  2. 2025

    Curse of dimensionality in Bayesian model updating

    Binbin Li, Zihan Liao

    arXiv preprint, arXiv:2506.17744

    arXiv

* Corresponding author.

See also my Google Scholar profile.

Open-source code

MATLAB

SeMIS for Reliability Estimation

Sequential multiple importance sampling for estimating small failure probabilities, including in high-dimensional problems, with optional single-run CoV bounds and twelve benchmark examples.

Mechanical Systems and Signal Processing, 2026

Sequential-multiple-importance-sampling-for-Reliability-Estimation Python

SeMIS for Reliability Estimation (Python)

Python port of the MATLAB SeMIS reliability code (NumPy/SciPy), with the same classes and twelve benchmark examples, checked against the MATLAB results.

Mechanical Systems and Signal Processing, 2026

Sequential-multiple-importance-sampling-for-Reliability-Estimation-Python MATLAB

SeMIS for Multimodal Bayesian Inference

Softly truncated, prior-based proposal sequences combined through multiple importance sampling for Bayesian evidence estimation and posterior sampling with multimodal, high-dimensional posteriors.

Mechanical Systems and Signal Processing, 2026

Sequential-multiple-importance-sampling-for-multimodal-Bayesian-inference Python

SeMIS for Multimodal Bayesian Inference (Python)

Python port of the MATLAB SeMIS code for evidence estimation and posterior sampling (NumPy/SciPy), with the same benchmarks, checked against the MATLAB results.

Mechanical Systems and Signal Processing, 2026

Sequential-multiple-importance-sampling-for-multimodal-Bayesian-inference-Python MATLAB

Relaxed Subset Simulation

Relaxed subset simulation with hybrid intermediate domains and the original sequential probability-product estimator for reliability estimation, with numerical examples A–C.

Reliability Engineering & System Safety, 2025

Relaxed-Subset-Simulation Python

Relaxed Subset Simulation (Python)

Python implementation of relaxed subset simulation with the original sequential probability-product estimator, with the same numerical examples A–C and corresponding source files.

Reliability Engineering & System Safety, 2025

Relaxed-Subset-Simulation-Python MATLAB

Generalized Subset Simulation with Multiple Importance Sampling (GSuS-MIS)

Generalized subset simulation with multiple importance sampling for reliability estimation with multiple failure events, with two numerical examples.

Computers & Structures, 2025

Generalized-Subset-Simulation-with-Multiple-Importance-Sampling Python

Generalized Subset Simulation with Multiple Importance Sampling (GSuS-MIS, Python)

Python implementation of generalized subset simulation with multiple importance sampling, with the same two numerical examples.

Computers & Structures, 2025

Generalized-Subset-Simulation-with-Multiple-Importance-Sampling-Python MATLAB

Subset Simulation for Bayesian Updating

Treats the likelihood as a limit-state function and uses subset simulation to estimate the Bayesian evidence, with a CoV correction for correlated Markov chain samples, and to draw posterior samples.

Mechanical Systems and Signal Processing, 2025

Subset-Simulation-for-Bayesian-Updating Python

Subset Simulation for Bayesian Updating (Python)

Python port of the MATLAB subset simulation code for Bayesian evidence estimation and posterior sampling (NumPy/SciPy), checked against the MATLAB results.

Mechanical Systems and Signal Processing, 2025

Subset-Simulation-for-Bayesian-Updating-Python MATLAB

1D-CNN-Assisted SERS Quantification

A one-dimensional convolutional neural network for quantitative regression of surface-enhanced Raman spectra, reporting calibration and test-set metrics including RMSEP and LOD.

Analytical Chemistry, 2026

1D-CNN-Assisted-SERS-Quantification Python

1D-CNN-Assisted SERS Quantification (Python)

PyTorch implementation of the MATLAB four-block 1D convolutional regression network for SERS quantification, with training, prediction, regression metrics and a self-contained synthetic-data demonstration.

Analytical Chemistry, 2026

1D-CNN-Assisted-SERS-Quantification-Python

Datasets

Open experimental datasets from structural vibration tests I participated in, published with my collaborators on Harvard Dataverse.