arXiv subjects

Lijing Wang

Publications and source records attributed to Lijing Wang.

13 recordsLinked to original sources

arXiv2026

CRAFT: LLM-Based Iterative Refinement for Temporal Reasoning over Clinical Narratives

Understanding the temporal progression of symptoms in clinical narratives is critical for disease monitoring, safety surveillance, and causality assessment. Clinical narratives, however, rarely provide explicit temporal anchors. Current approaches to temporal information reasoning focus predominantly on pairwise relation classification across multi-visit and timestamp-rich records, leaving the reconstruction of structured symptom trajectories from individual anchor-sparse reports largely unaddressed. We propose CRAFT, an LLM framework that pairs a generator with a constraint-based verifier to iteratively produce and refine stage-wise symptom timelines through targeted feedback. We conduct evaluation on MedTempo, a new benchmark of 5,347 vaccine adverse-event narratives spanning three COVID-19 vaccine types, with expert-validated temporal stage annotations for 3,166 reports. Experiments across four LLM backbones demonstrate that CRAFT consistently improves temporal ordering accuracy, with ablation analysis isolating the contribution of generator and verifier components across model capability levels.

cs.CL↗

arXiv2026

GRAIN: Group Aggregation via Min-Norm Objective

Learning instability is a long-standing problem across machine learning, but it is especially acute in the overparameterized regime that defines modern deep learning: large models fine-tuned or trained on limited data traverse flat loss landscapes with many nearly-equivalent minima, and stochastic factors (initialization, data order, dropout, hardware non-determinism) can route optimization to very different solutions. The rise of large pretrained models (LPMs) makes the problem more urgent: training cost is high, downstream data is often small, and repeated runs for variance reduction are prohibitive. We introduce \textbf{GRAIN} (\textbf{G}roup \textbf{A}ggregation via m\textbf{IN}-norm objective), a lightweight training algorithm that replaces the mean aggregation used in mini-batch optimization (both across mini-batches and within a mini-batch) with a min-norm convex combination of group-wise gradients. \mName guarantees a non-negative inner product between the aggregated update and every group gradient, resolving intra- and inner-batch gradient conflict, and retains an $\mathcal{O}(1/T)$ convergence rate comparable to SGD. Under mild smoothness and absolute-continuity assumptions, the min-norm solution differs almost surely from the arithmetic mean, which yields a uniform-stability bound for \mName strictly tighter than the standard bound for SGD. Empirically across generation, classification, and regression at LPM scale, \mName delivers consistent improvements in mean performance and reductions in run-to-run variance over a broad suite of tasks, with no extra training-time or storage cost beyond a single backward pass.

cs.LG↗

arXiv2026

Light-FMP: Lightweight Feature and Model Pruning for Enhanced Deep Recommender Systems

Deep recommender systems (DRS) often face challenges in balancing computational efficiency and model accuracy, especially when handling high-dimensional input features. Existing methods either focus on improving accuracy while neglecting training efficiency or prioritize efficiency at the cost of suboptimal accuracy across tasks. We propose Light-FMP: Lightweight Feature and Model Pruning for Enhanced DRS, a lightweight framework that addresses the challenges through three key phases: \textit{pretraining}, \textit{pruning}, and \textit{continued training}. Using a hard concrete distribution, a masking layer is efficiently pretrained on a small data subset to identify important features. The model and features are then pruned, and training continues on the remaining dataset with domain-adapted parameters. Experiments on benchmark datasets from real-world recommender systems demonstrate that Light-FMP outperforms existing methods in both efficiency and accuracy while maintaining scalability and robustness.

cs.IR↗

arXiv2026

Dynamic Scaled Gradient Descent for Stable Fine-Tuning for Classifications

Fine-tuning pretrained models has become a standard approach to adapting pretrained knowledge to improve the accuracy on new sparse, imbalance datasets. However, issues arise when optimization falls into a collapsed state, where the model gets stuck, leading to degraded performance and unstable training. One possible reason for this is the cancellation of gradients across training examples. To address this problem, we propose a novel algorithm, dynamic scaled gradient descent (\mName), that directly modifies the gradients returned by training examples, specifically, scaling down the gradients of correctly classified examples using a dynamic scaler. This strategy offers both theoretical and empirical advantages in improving training stability. Experiments on a variety of benchmark datasets, spanning multiple tasks and large pretrained models, demonstrate that our method consistently reduces performance variance and surpasses the accuracy of existing approaches.

cs.LG↗

arXiv2025

Assessing the Macro and Micro Effects of Random Seeds on Fine-Tuning Large Language Models

The impact of random seeds in fine-tuning large language models (LLMs) has been largely overlooked despite its potential influence on model performance.In this study, we systematically evaluate the effects of random seeds on LLMs using the GLUE and SuperGLUE benchmarks. We analyze the macro-level impact through traditional metrics like accuracy and F1, calculating their mean and variance to quantify performance fluctuations. To capture the micro-level effects, we introduce a novel metric, consistency, measuring the stability of individual predictions across runs. Our experiments reveal significant variance at both macro and micro levels, underscoring the need for careful consideration of random seeds in fine-tuning and evaluation.

cs.CL↗

arXiv2022

Multi-scale Digital Twin: Developing a fast and physics-informed surrogate model for groundwater contamination with uncertain climate models

Soil and groundwater contamination is a pervasive problem at thousands of locations across the world. Contaminated sites often require decades to remediate or to monitor natural attenuation. Climate change exacerbates the long-term site management problem because extreme precipitation and/or shifts in precipitation/evapotranspiration regimes could re-mobilize contaminants and proliferate affected groundwater. To quickly assess the spatiotemporal variations of groundwater contamination under uncertain climate disturbances, we developed a physics-informed machine learning surrogate model using U-Net enhanced Fourier Neural Operator (U-FNO) to solve Partial Differential Equations (PDEs) of groundwater flow and transport simulations at the site scale.We develop a combined loss function that includes both data-driven factors and physical boundary constraints at multiple spatiotemporal scales. Our U-FNOs can reliably predict the spatiotemporal variations of groundwater flow and contaminant transport properties from 1954 to 2100 with realistic climate projections. In parallel, we develop a convolutional autoencoder combined with online clustering to reduce the dimensionality of the vast historical and projected climate data by quantifying climatic region similarities across the United States. The ML-based unique climate clusters provide climate projections for the surrogate modeling and help return reliable future recharge rate projections immediately without querying large climate datasets. In all, this Multi-scale Digital Twin work can advance the field of environmental remediation under climate change.

physics.geo-ph↗

arXiv2022

A Derivation of Classical Orthogonal Polynomials using Generalized Vandermonde Determinants

We present a derivation of classical Hermite, Laguerre, and Jacobi orthogonal polynomials directly through the Gram-Schmidt orthogonization process. The derivation uses certain generalized Vandermonde determinants with entries defined by Gamma and Beta functions. We also provide a geometric formulation of Gram-Schmidt orthogonalization using the Hodge star operator.

math.RA↗

arXiv2020

Cohorting to isolate asymptomatic spreaders: An agent-based simulation study on the Mumbai Suburban Railway

The Mumbai Suburban Railways, \emph{locals}, are a key transit infrastructure of the city and is crucial for resuming normal economic activity. To reduce disease transmission, policymakers can enforce reduced crowding and mandate wearing of masks. \emph{Cohorting} -- forming groups of travelers that always travel together, is an additional policy to reduce disease transmission on \textit{locals} without severe restrictions. Cohorting allows us to: ($i$) form traveler bubbles, thereby decreasing the number of distinct interactions over time; ($ii$) potentially quarantine an entire cohort if a single case is detected, making contact tracing more efficient, and ($iii$) target cohorts for testing and early detection of symptomatic as well as asymptomatic cases. Studying impact of cohorts using compartmental models is challenging because of the ensuing representational complexity. Agent-based models provide a natural way to represent cohorts along with the representation of the cohort members with the larger social network. This paper describes a novel multi-scale agent-based model to study the impact of cohorting strategies on COVID-19 dynamics in Mumbai. We achieve this by modeling the Mumbai urban region using a detailed agent-based model comprising of 12.4 million agents. Individual cohorts and their inter-cohort interactions as they travel on locals are modeled using local mean field approximations. The resulting multi-scale model in conjunction with a detailed disease transmission and intervention simulator is used to assess various cohorting strategies. The results provide a quantitative trade-off between cohort size and its impact on disease dynamics and well being. The results show that cohorts can provide significant benefit in terms of reduced transmission without significantly impacting ridership and or economic \& social activity.

cs.MA↗

arXiv2020

Examining Deep Learning Models with Multiple Data Sources for COVID-19 Forecasting

The COVID-19 pandemic represents the most significant public health disaster since the 1918 influenza pandemic. During pandemics such as COVID-19, timely and reliable spatio-temporal forecasting of epidemic dynamics is crucial. Deep learning-based time series models for forecasting have recently gained popularity and have been successfully used for epidemic forecasting. Here we focus on the design and analysis of deep learning-based models for COVID-19 forecasting. We implement multiple recurrent neural network-based deep learning models and combine them using the stacking ensemble technique. In order to incorporate the effects of multiple factors in COVID-19 spread, we consider multiple sources such as COVID-19 confirmed and death case count data and testing data for better predictions. To overcome the sparsity of training data and to address the dynamic correlation of the disease, we propose clustering-based training for high-resolution forecasting. The methods help us to identify the similar trends of certain groups of regions due to various spatio-temporal effects. We examine the proposed method for forecasting weekly COVID-19 new confirmed cases at county-, state-, and country-level. A comprehensive comparison between different time series models in COVID-19 context is conducted and analyzed. The results show that simple deep learning models can achieve comparable or better performance when compared with more complicated models. We are currently integrating our methods as a part of our weekly forecasts that we provide state and federal authorities.

cs.LG↗

arXiv2020

Wisdom of the Ensemble: Improving Consistency of Deep Learning Models

Deep learning classifiers are assisting humans in making decisions and hence the user's trust in these models is of paramount importance. Trust is often a function of constant behavior. From an AI model perspective it means given the same input the user would expect the same output, especially for correct outputs, or in other words consistently correct outputs. This paper studies a model behavior in the context of periodic retraining of deployed models where the outputs from successive generations of the models might not agree on the correct labels assigned to the same input. We formally define consistency and correct-consistency of a learning model. We prove that consistency and correct-consistency of an ensemble learner is not less than the average consistency and correct-consistency of individual learners and correct-consistency can be improved with a probability by combining learners with accuracy not less than the average accuracy of ensemble component learners. To validate the theory using three datasets and two state-of-the-art deep learning classifiers we also propose an efficient dynamic snapshot ensemble method and demonstrate its value.

cs.LG↗

arXiv2020

TDEFSI: Theory Guided Deep Learning Based Epidemic Forecasting with Synthetic Information

Influenza-like illness (ILI) places a heavy social and economic burden on our society. Traditionally, ILI surveillance data is updated weekly and provided at a spatially coarse resolution. Producing timely and reliable high-resolution spatiotemporal forecasts for ILI is crucial for local preparedness and optimal interventions. We present TDEFSI (Theory Guided Deep Learning Based Epidemic Forecasting with Synthetic Information), an epidemic forecasting framework that integrates the strengths of deep neural networks and high-resolution simulations of epidemic processes over networks. TDEFSI yields accurate high-resolution spatiotemporal forecasts using low-resolution time series data. During the training phase, TDEFSI uses high-resolution simulations of epidemics that explicitly model spatial and social heterogeneity inherent in urban regions as one component of training data. We train a two-branch recurrent neural network model to take both within-season and between-season low-resolution observations as features, and output high-resolution detailed forecasts. The resulting forecasts are not just driven by observed data but also capture the intricate social, demographic and geographic attributes of specific urban regions and mathematical theories of disease propagation over networks. We focus on forecasting the incidence of ILI and evaluate TDEFSI's performance using synthetic and real-world testing datasets at the state and county levels in the USA. The results show that, at the state level, our method achieves comparable/better performance than several state-of-the-art methods. At the county level, TDEFSI outperforms the other methods. The proposed method can be applied to other infectious diseases as well.

stat.OT↗

arXiv2019

Graph Message Passing with Cross-location Attentions for Long-term ILI Prediction

Forecasting influenza-like illness (ILI) is of prime importance to epidemiologists and health-care providers. Early prediction of epidemic outbreaks plays a pivotal role in disease intervention and control. Most existing work has either limited long-term prediction performance or lacks a comprehensive ability to capture spatio-temporal dependencies in data. Accurate and early disease forecasting models would markedly improve both epidemic prevention and managing the onset of an epidemic. In this paper, we design a cross-location attention based graph neural network (Cola-GNN) for learning time series embeddings and location aware attentions. We propose a graph message passing framework to combine learned feature embeddings and an attention matrix to model disease propagation over time. We compare the proposed method with state-of-the-art statistical approaches and deep learning models on real-world epidemic-related datasets from United States and Japan. The proposed method shows strong predictive performance and leads to interpretable results for long-term epidemic predictions.

cs.LG↗

arXiv2019

Evaluating entropy rate of laser chaos and shot noise

Evaluating entropy rate of high-dimensional chaos and shot noise from analog raw signals remains elusive and important in information security. We experimentally present an accurate assessment of entropy rate for physical process randomness. The entropy generation of optical-feedback laser chaos and physical randomness limit from shot noise are quantified and unambiguously discriminated using the growth rate of average permutation entropy value in memory time. The permutation entropy difference of filtered laser chaos with varying embedding delay time is investigated experimentally and theoretically. High resolution maps of the entropy difference is observed over the range of the injection-feedback parameter space. We also clarify an inverse relationship between the entropy rate and time delay signature of laser chaos over a wide range of parameters. Compared to the original chaos, the time delay signature is suppressed up to 95% with the minimum of 0.015 via frequency-band extractor, and the experiment agrees well with the theory. Our system provides a commendable entropy evaluation and source for physical random number generation.

physics.optics↗