arXiv subjects

Yawen Shen

Publications and source records attributed to Yawen Shen.

3 recordsLinked to original sources

arXiv2026

Selectivity Matters: Source Node Influence Pruning for Unsupervised Graph Domain Adaptation

Unsupervised Graph Domain Adaptation (UGDA) aims to facilitate knowledge transfer from a labeled source graph to an unlabeled target graph by mitigating cross-domain distribution shifts. Existing methods primarily focus on node-level feature alignment in latent spaces, relying on the implicit assumption that all source nodes contribute positively to the alignment. However, this assumption often fails because a node's semantic information is intrinsically coupled with its topological graph structure. Due to structural shifts, source nodes with severe structural deviations (e.g., structural outliers) lack semantic counterparts in the target graph, and forcing alignment on them introduces severe noise and causes negative transfer. To bridge this gap, we argue that selective source node utilization is superior to full-graph training, thereby shifting the research paradigm from feature-level alignment to data-level refinement. To this end, we propose Source Node Influence Pruning (SNIP), a novel model-agnostic, data-centric refinement framework. Specifically, SNIP quantifies the structural discrepancy between individual source nodes and the target domain by integrating multiple centrality measures, assigning each source node an influence score. A rank-based normalization mechanism is further employed to eliminate scale variations across different measures, allowing SNIP to effectively identify and filter out structurally incompatible nodes with low influence scores. As a plug-and-play method, SNIP constructs a refined "sub-source" graph that is inherently more beneficial for subsequent alignment. Comprehensive experiments across eight transfer scenarios on five real-world datasets demonstrate that SNIP consistently outperforms competitive baselines and significantly enhances adaptation performance, validating the superiority of selective node utilization over full-graph training.

cs.LG↗

arXiv2025

Harvesting AlphaEarth: Benchmarking the Geospatial Foundation Model for Agricultural Downstream Tasks

Geospatial foundation models (GFMs) have emerged as a promising approach to overcoming the limitations in existing featurization methods. More recently, Google DeepMind has introduced AlphaEarth Foundation (AEF), a GFM pre-trained using multi-source EOs across continuous time. An annual and global embedding dataset is produced using AEF that is ready for analysis and modeling. The internal experiments show that AEF embeddings have outperformed operational models in 15 EO tasks without re-training. However, those experiments are mostly about land cover and land use classification. Applying AEF and other GFMs to agricultural monitoring require an in-depth evaluation in critical agricultural downstream tasks. There is also a lack of comprehensive comparison between the AEF-based models and traditional remote sensing (RS)-based models under different scenarios, which could offer valuable guidance for researchers and practitioners. This study addresses some of these gaps by evaluating AEF embeddings in three agricultural downstream tasks in the U.S., including crop yield prediction, tillage mapping, and cover crop mapping. Datasets are compiled from both public and private sources to comprehensively evaluate AEF embeddings across tasks at different scales and locations, and RS-based models are trained as comparison models. AEF-based models generally exhibit strong performance on all tasks and are competitive with purpose-built RS-based models in yield prediction and county-level tillage mapping when trained on local data. However, we also find several limitations in current AEF embeddings, such as limited spatial transferability compared to RS-based models, low interpretability, and limited time sensitivity. These limitations recommend caution when applying AEF embeddings in agriculture, where time sensitivity, generalizability, and interpretability is important.

cs.LG↗

arXiv2021

Modeling the Combined Impact of Rainfall and Storm Tide on Coastal Cities under a Changing Climate: Transportation Infrastructure Impacts in Norfolk, Virginia USA as a Case Study

Low-lying coastal cities across the world are vulnerable to the combined impact of rainfall and storm tide. However, existing approaches lack the ability to model the combined effect of these flood mechanisms. Thus, to increase flood resilience, modeling techniques to improve understanding and prediction of the combined effect of these flood hazards are critical. To address this need, this study presents a modeling system for assessing the combined flood risk to coastal cities under changing climate conditions that leverages ocean modeling with land surface modeling capable of resolving urban drainage infrastructure within the city. The modeling approach is demonstrated in quantifying the future impact on transportation infrastructure within Norfolk, Virginia USA. A series of combined storms events are modeled for current (2020) and projected future (2070) climate conditions. Results show that pluvial flooding causes a larger interruption to the transportation network compared to tidal flooding under current climate conditions. By 2070, however, tidal flooding will be the dominant flooding mechanism with even nuisance flooding expected to happen daily due to SLR. In 2070, nuisance flooding is expected to cause a 4.6% total link close time (TLC), which is more than two times that of a 50-year storm surge (1.8% TLC) in 2020. The coupled model was compared with a widely used but physically simplistic bathtub method to assess the difference resulting from the more complex modeling presented. Results show that the bathtub method overestimated the flooded area near the shoreline by 9.5% and 3.1% for a 10-year storm surge event in 2020 and 2070, respectively, but underestimated flooded area in the inland region by 9.0% and 4.0% for the same events. The findings demonstrate the benefit of sophisticated modeling methods in climate adaptive planning and policy in coastal communities.

physics.ao-ph↗