Jiguang Li, Robert Gibbons, and Veronika Ročková.
"Sparse Bayesian Multidimensional Item Response Theory."
Journal of the American Statistical Association, 2025 (in press).
[Paper]
[Code]
Jiguang Li, Robert Gibbons, and Veronika Ročková.
"Deep Computerized Adaptive Testing." Submitted, 2025.
[Paper]
[Code]
Past Projects
Astrostatistics: Developed a Python implementation of the Alpha-shape Fitting to Spectrum (AFS) and
the Alpha-shape and Lab Source Fitting to Spectrum (ALSFS) algorithms for spectrum continuum flattening,
a crucial preprocessing step in spectroscopic analysis.
[Code].
Algorithms originally proposed in Xin Xu et al. (2019).
Deep Learning for Medical Imaging: Implemented custom DenseNet and ResNet architectures and
applied ensemble learning methods to enhance X-ray fracture prediction.
Achieved 0.92 AUC and 0.75 Kappa on the Stanford MURA dataset,
outperforming the baseline model (0.705 Kappa) from the
original 2018 MURA paper.
Sample saliency maps available here.
Optimal Transport: Expository writing on robust estimation of Wasserstein distance via factored couplings.
[Paper].
Mixing Markov Chains: Investigated convex optimization methods for the fastest mixing Markov chain.
Applied weak duality to derive the optimal transition probability matrix for star graphs.
[Paper].
Undergraduate Thesis (Algebraic Combinatorics): The Chevalley-Warning Theorem: Its Proofs,
Generalizations, and Applications.
[Paper].
3D Reconstruction: Undergraduate computer science project on 3D shape reconstruction from images.
[Paper].