Global ETD Search
Search theses and dissertations gathered from participating repositories worldwide. Every result links back to the library that holds it. No account is needed.
Results
Showing 1 to 19 of 19 for “"Oversmoothing"”.
-
Efficient approaches to deal with oversmoothing in deep graph neural networks
… depth increases due to the phenomenon known as oversmoothing, where node representations become indistinguishable across layers. This thesis presents a systematic investigation into the oversmoothing problem and proposes novel theoretical and empirical approaches for mitigating it, thus enabling …
-
Investigation of tail probability using a smooth estimation of a survival function
… their smooth estimator. Nevertheless, in fear of oversmoothing with the untruncated version, the authors did not pursue research on this estimator much further. This thesis deals with the investigation of some statistical properties of the untruncated version of the smooth estimator and studies …
-
Human motion prediction.
… models suffer from the mean pose problem, and oversmoothing problem. Mean pose problem is the models prone to produce the fix pose for a time period. The oversmoothing problem is spefic exist in the graph neural network models for motion prediction. It means that when the neural network goes …
-
Spectral-spatial classification techniques for hyperspectral imagery
… parameters used. This method can overcome the oversmoothing effect of the first technique. Thirdly, an extension of the second approach is developed in this study based on a multi-scale conservative smoothing scheme and adaptive sparse representation. This method can automatically overcome the …
-
Regularization in microwave tomography
… investigated to obtain a stable solution without oversmoothing the solution. Tikhonov and Edge-preserving regularizations are compared. It is found that edge-preserving regularization decreases the smoothness of the reconstruction but has the same robustness against the noise compared to Tikhonov …
-
Theoretical foundations of Graph Neural Networks: from expressivity to generalization
… than 1-WL. A key expressivity limitation is oversmoothing, where node embeddings become indistinguishable as the network depth increases. To address this, we develop a GNN architecture based on fractional graph Laplacians embedded into neural ODEs (Maskey et al., 2023b). This approach enables …
-
Multi-View Pedestrian Detection Without Scene-Specific Fine-Tuning via 3D Foundation Models
… mechanisms like soft-depth sampling can cause oversmoothing, and cross-subset experiments reveal that BEV convolutional networks severely overfit to the training camera rig layout (dropping up to 28.35 MODA when testing on unseen camera subsets).
-
Contributions to regularization theory and practice of certain nonlinear inverse problems
… operator equations in Hilbert scales with oversmoothing penalties. Sufficient as well as necessary conditions to establish convergence are introduced and convergence rate results are given for various parameter choice rules under a two sided nonlinearity constraint. Ultimately, both a …
-
A conservative prior for bayesian hierarchical models in biostatistics
… is more serious type of error than oversmoothing. So, we need an estimation approach which can guard against undersmoothing. If we can control the undersmoothing reasonably, we may get a better calibrated summary of the phenomenon we estimate. In this study, we have incorporated the …
-
Graph-Based Learning: From Graph Neural Networks to Large Language Models
… First, we introduce LSGAT, which mitigates oversmoothing in deep GAT architectures by dynamically adjusting attention across layers. Second, we propose LangGSL, a variational mutual learning framework that couples LLMs with graph structure learning models, enabling robust representation …
-
Geometric Deep Learning for High Dimensional Inference and Sampling
… graph neural networks, addressing the issues of oversmoothing and heterophily commonly encountered in graph learning. In the second part, we focus on uncertainty quantification in geometry. We begin with the protein inverse folding problem, a discrete data sampling problem where each variable is …
-
Graph Representation Learning for Classification and Anomaly Detection
… model addresses the challenges of oversmoothing and shrinking effects by introducing a nonlinear smoothness term into the feature diffusion mechanism of graph convolutional networks. We conduct comprehensive experiments on diverse benchmark datasets demonstrating that our approach …
-
A Neuro-Symbolic Framework to Enhance Interpretability of Graph Neural Networks
… This analysis is then extended to examine depth, oversmoothing, and their effect on discriminative power. Building on the same activation extraction method, we introduce a semantic activation-mining approach that extracts compact rule-based explanations from binarised hidden activations, producing …
-
The application of the jackknife in geostatistical resource estimation: Robust estimator and its measure of uncertainty.
… at the same time reduces its local bias and oversmoothing tendency. The second contribution is the ability, through the jackknife standard deviation, to set a confidence limit for a reserve estimate of a general shape. Jackknifing the ordinary kriging estimate maximizes sample utilization, as …
-
Pose Estimation and Object Detection using Deep Convolutional Networks
… joints. Moreover, our approach alleviates the oversmoothing problem caused by repeated graph convolutions, preventing the learned feature representations from converging to similar values thanks in part to residual connections with the first layer of the network. These residual connections are …
-
Semiparametric Estimation of Unimodal Distributions
… degree of smoothing. Problems such as under and oversmoothing must be addressed in order to determine the shape of the underlying distribution. Furthermore, smoothing of skewed data can give a biased estimation of the peak position. We propose two new approaches for statistical mode estimation …
-
Automatic detection of drusen associated with age-related macular degeneration in optical coherence tomography: a graph-based approach
… and regularization can still cause the oversmoothing of some drusen. To address this problem, it is also proposed the integration of local shape priors in the form of a sparse high order potentials (SHOPs) into the multiple surface framework. Drusen are commonly detected by thresholding …
-
Geostatistical methods for estimating iron, silica and alumina grades within the hardcap of the section seven iron deposit, Tom Price
… estimate indicates that there is a risk of oversmoothing with regard to the HG cut-off then Median IK should be used to identify areas which have a chance of being HG and then deciding on the best way to take advantage of this information. Some possibilities would be to: target these areas …
-
Deep Learning-Based Respiratory Motion Correction for Positron Emission Tomography
… deformations, it does sometimes lead to oversmoothing effects or distortions in non-moving structures, particularly in low-contrast or highly motion-affected regions. Overall, this work demonstrated that the proposed unsupervised deep learning-based image registration framework can be an …