arXiv subjects

Jialu Sun

Publications and source records attributed to Jialu Sun.

3 recordsLinked to original sources

arXiv2026

Uncertainty aware prediction from incomplete clinical records: a conformal prediction study of Parkinsonian classification using C-OPN data

Background: Incomplete clinical records and imbalanced diagnostic groups complicate Parkinsonian classification. We examined whether incomplete records could support useful lists of diagnostic possibilities through conformal prediction, assessing how often the lists retained the recorded diagnosis, their length and the uncertainty remaining. Methods: We analyzed records from 3,434 Canadian Open Parkinson Network participants: 2,853 with Parkinson's disease, 171 in the atypical Parkinsonian group (AP) and 410 healthy controls. In three repeats of five-fold cross-validation, we compared single-category predictions, conformal prediction sets and lists combined across 21 analyses using alternative models and assumptions about missing entries. An information deficit score summarized unavailable information. Results: The median proportion of eligible predictor entries unrecorded per participant was 75.6%. Single-category prediction correctly classified 89.29% of test records (balanced accuracy 55.40%), but only 10.33% of AP records. Conformal prediction sets included the recorded category in 93.35% of evaluation records and averaged 1.846 categories; combined lists increased inclusion to approximately 99% and averaged 2.289 categories. AP inclusion increased from 98.05% to 100%. Combined lists averaged 1.010 more categories in the group with the greatest information deficit than in the group with the least (95% interval 0.996-1.023). Conclusions: Even with extensive missing information and imbalanced diagnostic groups, clinical records can support informative prediction when uncertainty is quantified and expressed through sets of possible diagnoses rather than a single label. Joint assessment of list length and information deficit showed both the narrowing achieved and the records for which all three categories remained open.

stat.AP↗

arXiv2026

Open-Source Autonomous Driving System Analysis and Multi-Disciplinary Hardware-in-the-Loop Research Paradigm with Reinforcement-Learning Testing and Large Language Models

Open-source autonomous driving systems provide an inspectable software foundation for intelligent vehicle research. Under real-vehicle deployment conditions, the recording and review of experimental conditions are important for interpreting system behavior and reusing experimental results. However, in a shared real-vehicle environment involving multiple vehicles, task processes, code modifications, and hardware testing feedback are often distributed across different teams and experimental stages, making it challenging to maintain continuous and reviewable experimental records. To address this limitation, this paper examines an Apollo-on-Hongqi EV environment and proposes a real-vehicle experimental framework. The framework connects multi-vehicle experiments, repository-based code reuse and software-hardware testing feedback within a unified review process. Large language models and RL-based testing serve as auxiliary components for record organization, anomaly summarization, and simulation-based candidate scenario generation. Based on this setting, this paper analyzes preliminary evidence from multi-vehicle collaborative experimentation, code and experimental-skill sharing, and software-hardware collaborative testing. The analysis shows that experimental records can be examined together with their operating conditions, providing a reviewable basis for Apollo-on-Hongqi EV research.

cs.SE↗

arXiv2026

Improving Graph Few-shot Learning with Hyperbolic Space and Denoising Diffusion

Graph few-shot learning, which focuses on effectively learning from only a small number of labeled nodes to quickly adapt to new tasks, has garnered significant research attention. Despite recent advances in graph few-shot learning that have demonstrated promising performance, existing methods still suffer from several key limitations. First, during the meta-training phase, these methods typically perform node representation learning in Euclidean space, which often fails to capture the inherently hierarchical structure existing in real-world graph data. Second, during the meta-testing phase, they usually fit an empirical target distribution derived from only a few support samples, even when this distribution significantly deviates from the true underlying distribution. To address these issues, we propose IMPRESS, a novel framework that IMproves graPh few-shot learning with hypeRbolic spacE and denoiSing diffuSion. Specifically, our model learns node representations in a hyperbolic space and enriches the support distribution through denoising diffusion mechanisms. Theoretically, IMPRESS achieves a tighter generalization bound. Empirically, IMPRESS consistently outperforms competitive baselines across multiple benchmark datasets.

cs.LG↗