arXiv subjects

Harlin Lee

Publications and source records attributed to Harlin Lee.

At least 19 recordsLinked to original sources

arXiv2026

Uncovering Trajectory and Topological Signatures in Multimodal Pediatric Sleep Embeddings

While generative models have shown promise in pediatric sleep analysis, the latent structure of their multimodal embeddings remains poorly understood. This work investigates session-wide diagnostic information contained in the sequences of 30-second pediatric PSG epochs embedded by a multimodal masked autoencoder. We test whether augmenting embeddings with PHATE-derived per-epoch coordinates and whole-night movement descriptors, persistent homology summaries of the embedding cloud, and EHR yields task-relevant signals. Simple linear and MLP models, chosen for interpretability rather than state-of-the-art performance, show that geometric, topological, and clinical features each provide complementary gains. For binary predictions, feature importance is task-dependent, and more expressive late-fusion models generally perform better, with AUPRC improving from 0.26 to 0.34 for desaturation, 0.31 to 0.48 for EEG arousal, 0.09 to 0.22 for hypopnea, and 0.05 to 0.14 for apnea. We also report Brier score and Expected Calibration Error, where the full fusion model yields the best calibration across all four binary tasks. Our study reveals that latent geometry/topology and EHR offer complementary, interpretable signals beyond embeddings, improving calibration and robustness under extreme imbalance.

cs.LG↗

arXiv2026

Fused Gromov-Wasserstein Distance with Feature Selection

Fused Gromov-Wasserstein (FGW) distances provide a principled framework for comparing objects by jointly aligning structure and node features. However, existing FGW formulations treat all features uniformly, which limits interpretability and robustness in high-dimensional settings where many features may be irrelevant or noisy. We introduce FGW distances with feature selection, which incorporate adaptive feature suppression weights into the FGW objective to selectively downweight or suppress differentiating features during alignment. We propose two approaches: (1) regularized FGW with Lasso and Ridge penalties, and (2) FGW with simplex-constrained weights, including groupwise extensions. We analyze the resulting models and establish their key theoretical properties, including bounds relative to classical FGW and Gromov-Wasserstein distances, and metric behavior. An efficient alternating minimization algorithm is developed. Experiments illustrate how feature suppression enhances interpretability and reveals task-relevant structure, with a special application to computational redistricting.

cs.LG↗

arXiv2026

BiTimeCrossNet: Time-Aware Self-Supervised Learning for Pediatric Sleep

We present BiTimeCrossNet (BTCNet), a multimodal self-supervised learning framework for long physiological recordings such as overnight sleep studies. While many existing approaches train on short segments treated as independent samples, BTCNet incorporates information about when each segment occurs within its parent recording, for example within a sleep session. BTCNet further learns pairwise interactions between physiological signals via cross-attention, without requiring task labels or sequence-level supervision. We evaluate BTCNet on pediatric sleep data across six downstream tasks, including sleep staging, arousal detection, and respiratory event detection. Under frozen-backbone linear probing, BTCNet consistently outperforms an otherwise identical non-time-aware variant, with gains that generalize to an independent pediatric dataset. Compared to existing multimodal self-supervised sleep models, BTCNet achieves strong performance, particularly on respiration-related tasks.

cs.LG↗

arXiv2025

Scaling Unsupervised Multi-Source Federated Domain Adaptation through Group-Wise Discrepancy Minimization

Unsupervised multi-source domain adaptation (UMDA) leverages labeled data from multiple source domains to generalize to an unlabeled target. While federated UMDA addresses privacy by avoiding raw data sharing, existing methods scale poorly as the number of sources increases, often suffering from high computational overhead or training instability. We propose GALA, a scalable and robust federated UMDA framework designed for high-diversity settings. GALA achieves scalability by coupling a novel inter-group discrepancy minimization objective that approximates pairwise alignment with linear complexity alongside a temperature-controlled, centroid-based weighting strategy for dynamic source prioritization. These components enable stable, parallelizable training across many heterogeneous sources, addressing a critical scalability bottleneck that remains largely unaddressed in current literature. To evaluate performance in high-diversity scenarios, we introduce Digit-18, a new benchmark comprising 18 datasets with varied synthetic and real-world domain shifts. Extensive experiments demonstrate that GALA achieves state-of-the-art results on standard benchmarks and significantly outperforms prior methods in large-scale settings where others either fail to converge or become computationally infeasible.

cs.LG↗

arXiv2025

RELATE: Relation Extraction in Biomedical Abstracts with LLMs and Ontology Constraints

Biomedical knowledge graphs (KGs) are vital for drug discovery and clinical decision support but remain incomplete. Large language models (LLMs) excel at extracting biomedical relations, yet their outputs lack standardization and alignment with ontologies, limiting KG integration. We introduce RELATE, a three-stage pipeline that maps LLM-extracted relations to standardized ontology predicates using ChemProt and the Biolink Model. The pipeline includes: (1) ontology preprocessing with predicate embeddings, (2) similarity-based retrieval enhanced with SapBERT, and (3) LLM-based reranking with explicit negation handling. This approach transforms relation extraction from free-text outputs to structured, ontology-constrained representations. On the ChemProt benchmark, RELATE achieves 52% exact match and 94% accuracy@10, and in 2,400 HEAL Project abstracts, it effectively rejects irrelevant associations (0.4%) and identifies negated assertions. RELATE captures nuanced biomedical relationships while ensuring quality for KG augmentation. By combining vector search with contextual LLM reasoning, RELATE provides a scalable, semantically accurate framework for converting unstructured biomedical literature into standardized KGs.

cs.IR↗

arXiv2025

A Closeness Centrality-based Circuit Partitioner for Quantum Simulations

Simulating quantum circuits (QC) on high-performance computing (HPC) systems has become an essential method to benchmark algorithms and probe the potential of large-scale quantum computation despite the limitations of current quantum hardware. However, these simulations often require large amounts of resources, necessitating the use of large clusters with thousands of compute nodes and large memory footprints. In this work, we introduce an end-to-end framework that provides an efficient partitioning scheme for large-scale QCs alongside a flexible code generator to offer a portable solution that minimizes data movement between compute nodes. By formulating the distribution of quantum states and circuits as a graph problem, we apply closeness centrality to assess gate importance and design a fast, scalable partitioning method. The resulting partitions are compiled into highly optimized codes that run seamlessly on a wide range of supercomputers, providing critical insights into the performance and scalability of quantum algorithm simulations.

quant-ph↗

arXiv2025

Federated Learning for Epileptic Seizure Prediction Across Heterogeneous EEG Datasets

Developing accurate and generalizable epileptic seizure prediction models from electroencephalography (EEG) data across multiple clinical sites is hindered by patient privacy regulations and significant data heterogeneity (non-IID characteristics). Federated Learning (FL) offers a privacy-preserving framework for collaborative training, but standard aggregation methods like Federated Averaging (FedAvg) can be biased by dominant datasets in heterogeneous settings. This paper investigates FL for seizure prediction using a single EEG channel across four diverse public datasets (Siena, CHB-MIT, Helsinki, NCH), representing distinct patient populations (adult, pediatric, neonate) and recording conditions. We implement privacy-preserving global normalization and propose a Random Subset Aggregation strategy, where each client trains on a fixed-size random subset of its data per round, ensuring equal contribution during aggregation. Our results show that locally trained models fail to generalize across sites, and standard weighted FedAvg yields highly skewed performance (e.g., 89.0% accuracy on CHB-MIT but only 50.8% on Helsinki and 50.6% on NCH). In contrast, Random Subset Aggregation significantly improves performance on under-represented clients (accuracy increases to 81.7% on Helsinki and 68.7% on NCH) and achieves a superior macro-average accuracy of 77.1% and pooled accuracy of 80.0% across all sites, demonstrating a more robust and fair global model. This work highlights the potential of balanced FL approaches for building effective and generalizable seizure prediction systems in realistic, heterogeneous multi-hospital environments while respecting data privacy.

cs.LG↗

arXiv2025

Adaptive Multimodal Protein Plug-and-Play with Diffusion-Based Priors

In an inverse problem, the goal is to recover an unknown parameter (e.g., an image) that has typically undergone some lossy or noisy transformation during measurement. Recently, deep generative models, particularly diffusion models, have emerged as powerful priors for protein structure generation. However, integrating noisy experimental data from multiple sources to guide these models remains a significant challenge. Existing methods often require precise knowledge of experimental noise levels and manually tuned weights for each data modality. In this work, we introduce Adam-PnP, a Plug-and-Play framework that guides a pre-trained protein diffusion model using gradients from multiple, heterogeneous experimental sources. Our framework features an adaptive noise estimation scheme and a dynamic modality weighting mechanism integrated into the diffusion process, which reduce the need for manual hyperparameter tuning. Experiments on complex reconstruction tasks demonstrate significantly improved accuracy using Adam-PnP.

cs.LG↗

arXiv2025

Empirical Power Analysis of a Statistical Test to Quantify Gerrymandering

Gerrymandering is a pervasive problem within the US political system. In the past decade, methods based on Markov Chain Monte Carlo (MCMC) sampling and statistical outlier tests have been proposed to quantify gerrymandering and were used as evidence in several high-profile legal cases. We perform an empirical power analysis of one such hypothesis test from Chikina et al (2020). We generate a family of biased North Carolina congressional district maps using the 2012 and 2016 presidential elections and assess under which conditions the outlier test fails to flag them at the specified Type I error level. The power of the outlier test is found to be relatively stable across political parties, election years, lengths of the MCMC chain and effect sizes. The main effect on the power of the test is shown to be the choice of the bias metric. This is the first work that computationally verifies the power of statistical tests used in gerrymandering cases.

cs.CY↗

arXiv2024

PedSleepMAE: Generative Model for Multimodal Pediatric Sleep Signals

Pediatric sleep is an important but often overlooked area in health informatics. We present PedSleepMAE, a generative model that fully leverages multimodal pediatric sleep signals including multichannel EEGs, respiratory signals, EOGs and EMG. This masked autoencoder-based model performs comparably to supervised learning models in sleep scoring and in the detection of apnea, hypopnea, EEG arousal and oxygen desaturation. Its embeddings are also shown to capture subtle differences in sleep signals coming from a rare genetic disorder. Furthermore, PedSleepMAE generates realistic signals that can be used for sleep segment retrieval, outlier detection, and missing channel imputation. This is the first general-purpose generative model trained on multiple types of pediatric sleep signals.

cs.LG↗

arXiv2024

Explainable Enrichment-Driven GrAph Reasoner (EDGAR) for Large Knowledge Graphs with Applications in Drug Repurposing

Knowledge graphs (KGs) represent connections and relationships between real-world entities. We propose a link prediction framework for KGs named Enrichment-Driven GrAph Reasoner (EDGAR), which infers new edges by mining entity-local rules. This approach leverages enrichment analysis, a well-established statistical method used to identify mechanisms common to sets of differentially expressed genes. EDGAR's inference results are inherently explainable and rankable, with p-values indicating the statistical significance of each enrichment-based rule. We demonstrate the framework's effectiveness on a large-scale biomedical KG, ROBOKOP, focusing on drug repurposing for Alzheimer disease (AD) as a case study. Initially, we extracted 14 known drugs from the KG and identified 20 contextual biomarkers through enrichment analysis, revealing functional pathways relevant to shared drug efficacy for AD. Subsequently, using the top 1000 enrichment results, our system identified 1246 additional drug candidates for AD treatment. The top 10 candidates were validated using evidence from medical literature. EDGAR is deployed within ROBOKOP, complete with a web user interface. This is the first study to apply enrichment analysis to large graph completion and drug repurposing.

cs.IT↗

arXiv2024

Efficient Trajectory Inference in Wasserstein Space Using Consecutive Averaging

Capturing data from dynamic processes through cross-sectional measurements is seen in many fields, such as computational biology. Trajectory inference deals with the challenge of reconstructing continuous processes from such observations. In this work, we propose methods for B-spline approximation and interpolation of point clouds through consecutive averaging that is intrinsic to the Wasserstein space. Combining subdivision schemes with optimal transport-based geodesic, our methods carry out trajectory inference at a chosen level of precision and smoothness, and can automatically handle scenarios where particles undergo division over time. We prove linear convergence rates and rigorously evaluate our method on cell data characterized by bifurcations, merges, and trajectory splitting scenarios like $supercells$, comparing its performance against state-of-the-art trajectory inference and interpolation methods. The results not only underscore the effectiveness of our method in inferring trajectories but also highlight the benefit of performing interpolation and approximation that respect the inherent geometric properties of the data.

cs.LG↗

arXiv2024

A Blob Method for Mean Field Control With Terminal Constraints

In the present work, we develop a novel particle method for a general class of mean field control problems, with source and terminal constraints. Specific examples of the problems we consider include the dynamic formulation of the p-Wasserstein metric, optimal transport around an obstacle, and measure transport subject to acceleration controls. Unlike existing numerical approaches, our particle method is meshfree and does not require global knowledge of an underlying cost function or of the terminal constraint. A key feature of our approach is a novel way of enforcing the terminal constraint via a soft, nonlocal approximation, inspired by recent work on blob methods for diffusion equations. We prove convergence of our particle approximation to solutions of the continuum mean-field control problem in the sense of Gamma-convergence. A byproduct of our result is an extension of existing discrete-to-continuum convergence results for mean field control problems to more general state and measure costs, as arise when modeling transport around obstacles, and more general constraint sets, including controllable linear time invariant systems. Finally, we conclude by implementing our method numerically and using it to compute solutions the example problems discussed above. We conduct a detailed numerical investigation of the convergence properties of our method, as well as its behavior in sampling applications and for approximation of optimal transport maps.

math.OC↗

arXiv2023

$O(k)$-Equivariant Dimensionality Reduction on Stiefel Manifolds

Many real-world datasets live on high-dimensional Stiefel and Grassmannian manifolds, $V_k(\mathbb{R}^N)$ and $Gr(k, \mathbb{R}^N)$ respectively, and benefit from projection onto lower-dimensional Stiefel and Grassmannian manifolds. In this work, we propose an algorithm called \textit{Principal Stiefel Coordinates (PSC)} to reduce data dimensionality from $ V_k(\mathbb{R}^N)$ to $V_k(\mathbb{R}^n)$ in an \textit{$O(k)$-equivariant} manner ($k \leq n \ll N$). We begin by observing that each element $\alpha \in V_n(\mathbb{R}^N)$ defines an isometric embedding of $V_k(\mathbb{R}^n)$ into $V_k(\mathbb{R}^N)$. Next, we describe two ways of finding a suitable embedding map $\alpha$: one via an extension of principal component analysis ($\alpha_{PCA}$), and one that further minimizes data fit error using gradient descent ($\alpha_{GD}$). Then, we define a continuous and $O(k)$-equivariant map $\pi_\alpha$ that acts as a "closest point operator" to project the data onto the image of $V_k(\mathbb{R}^n)$ in $V_k(\mathbb{R}^N)$ under the embedding determined by $\alpha$, while minimizing distortion. Because this dimensionality reduction is $O(k)$-equivariant, these results extend to Grassmannian manifolds as well. Lastly, we show that $\pi_{\alpha_{PCA}}$ globally minimizes projection error in a noiseless setting, while $\pi_{\alpha_{GD}}$ achieves a meaningfully different and improved outcome when the data does not lie exactly on the image of a linearly embedded lower-dimensional Stiefel manifold as above. Multiple numerical experiments using synthetic and real-world data are performed.

cs.CG↗

arXiv2022

Active Learning of Non-semantic Speech Tasks with Pretrained Models

Pretraining neural networks with massive unlabeled datasets has become popular as it equips the deep models with a better prior to solve downstream tasks. However, this approach generally assumes that the downstream tasks have access to annotated data of sufficient size. In this work, we propose ALOE, a novel system for improving the data- and label-efficiency of non-semantic speech tasks with active learning. ALOE uses pretrained models in conjunction with active learning to label data incrementally and learn classifiers for downstream tasks, thereby mitigating the need to acquire labeled data beforehand. We demonstrate the effectiveness of ALOE on a wide range of tasks, uncertainty-based acquisition functions, and model architectures. Training a linear classifier on top of a frozen encoder with ALOE is shown to achieve performance similar to several baselines that utilize the entire labeled data.

cs.SD↗

arXiv2022

Automatic Sleep Scoring from Large-scale Multi-channel Pediatric EEG

Sleep is particularly important to the health of infants, children, and adolescents, and sleep scoring is the first step to accurate diagnosis and treatment of potentially life-threatening conditions. But pediatric sleep is severely under-researched compared to adult sleep in the context of machine learning for health, and sleep scoring algorithms developed for adults usually perform poorly on infants. Here, we present the first automated sleep scoring results on a recent large-scale pediatric sleep study dataset that was collected during standard clinical care. We develop a transformer-based model that learns to classify five sleep stages from millions of multi-channel electroencephalogram (EEG) sleep epochs with 78% overall accuracy. Further, we conduct an in-depth analysis of the model performance based on patient demographics and EEG channels. The results point to the growing need for machine learning research on pediatric sleep.

eess.SP↗

arXiv2022

Predicting the Future of AI with AI: High-quality link prediction in an exponentially growing knowledge network

A tool that could suggest new personalized research directions and ideas by taking insights from the scientific literature could significantly accelerate the progress of science. A field that might benefit from such an approach is artificial intelligence (AI) research, where the number of scientific publications has been growing exponentially over the last years, making it challenging for human researchers to keep track of the progress. Here, we use AI techniques to predict the future research directions of AI itself. We develop a new graph-based benchmark based on real-world data -- the Science4Cast benchmark, which aims to predict the future state of an evolving semantic network of AI. For that, we use more than 100,000 research papers and build up a knowledge network with more than 64,000 concept nodes. We then present ten diverse methods to tackle this task, ranging from pure statistical to pure learning methods. Surprisingly, the most powerful methods use a carefully curated set of network features, rather than an end-to-end AI approach. It indicates a great potential that can be unleashed for purely ML approaches without human knowledge. Ultimately, better predictions of new future research directions will be a crucial component of more advanced research suggestion tools.

cs.AI↗

arXiv2022

Distilled Non-Semantic Speech Embeddings with Binary Neural Networks for Low-Resource Devices

This work introduces BRILLsson, a novel binary neural network-based representation learning model for a broad range of non-semantic speech tasks. We train the model with knowledge distillation from a large and real-valued TRILLsson model with only a fraction of the dataset used to train TRILLsson. The resulting BRILLsson models are only 2MB in size with a latency less than 8ms, making them suitable for deployment in low-resource devices such as wearables. We evaluate BRILLsson on eight benchmark tasks (including but not limited to spoken language identification, emotion recognition, health condition diagnosis, and keyword spotting), and demonstrate that our proposed ultra-light and low-latency models perform as well as large-scale models.

cs.SD↗