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 20 of 49 for “"maximum mean discrepancy"”.

  1. Image Analysis Applications of the Maximum Mean Discrepancy Distance Measure

    … signal processing world. The difference of group means computed using the Euclidean, or L2 distance, is one of the predominant distance measures used to compare feature vectors and groups of vectors, but many problems arise with it when high data dimensionality is present. Maximum mean discrepancy …

    waterloo Repository record for Image Analysis Applications of the Maximum Mean Discrepancy Distance Measure (opens in a new tab)

  2. Data-Adaptive Kernel Methods for Distributional Inference

    Kernel-based discrepancy measures, especially maximum mean discrepancy, provide flexible nonparametric tools for comparing probability distributions. Their practical performance, however, depends critically on the choice of kernel. In many applications, the kernel is selected or optimized using …

    gatech Repository record for Data-Adaptive Kernel Methods for Distributional Inference (opens in a new tab)

  3. Distance and Kernel-Based Nonparametric Two-Sample Tests

    … sizes both diverge to infinity. We focus on the maximum mean discrepancy (MMD) using isotropic kernel, including MMD with the Gaussian kernel and the Laplace kernel, and the energy distance as special cases. We establish the central limit theorem (CLT) under both the null hypothesis and the local …

    tamu Repository record for Distance and Kernel-Based Nonparametric Two-Sample Tests (opens in a new tab)

  4. Distribution distance measures in generative and privacy models

    … and increases the model's ability to generalize. Maximum mean discrepancy and energy distance are two such metrics that are easily defined and implemented over samples, and provide meaningful results on a range of data sets and data types. This work presents three main contributions: (1) a novel …

    texas Repository record for Distribution distance measures in generative and privacy models (opens in a new tab)

  5. TESTING EQUALITY OF DISTRIBUTIONS IN HIGH-DIMENSION:MAXIMUM MEAN DISCREPANCY BASED APPROACHES

    … area. However, the theoretical results of maximum mean discrepancy (MMD)-based tests in the literature are based on a restrictive equal sample size assumption. Therefore, a further analysis on these MMD-based tests for two-sample equal-distribution testing problems with this assumption …

    nus Repository record for TESTING EQUALITY OF DISTRIBUTIONS IN HIGH-DIMENSION:MAXIMUM MEAN DISCREPANCY BASED APPROACHES (opens in a new tab) Download file for TESTING EQUALITY OF DISTRIBUTIONS IN HIGH-DIMENSION:MAXIMUM MEAN DISCREPANCY BASED APPROACHES (opens in a new tab)

  6. A kernel-based approach to differentially private image generation

    … generator training scheme based on Maximum Mean Discrepancy (MMD) to mitigate privacy loss. Experimental results reveal the efficacy of our method. it successfully generates samples for CIFAR-10, CelebA, MNIST, and FashionMNIST. Theoretical analysis of our privacy-preserving loss …

    ubc Repository record for A kernel-based approach to differentially private image generation (opens in a new tab) Download PDF for A kernel-based approach to differentially private image generation (opens in a new tab)

  7. Variational Problems in Feature Selection and Data Compression

    … kernel, reducing local compression to maximum mean discrepancy minimization and kernel quadrature. For the Sinkhorn divergence, this geometry is equivalent to a scaled Gaussian reproducing kernel Hilbert space and supports coresets with poly-logarithmically many observations. The final …

    washington Repository record for Variational Problems in Feature Selection and Data Compression (opens in a new tab)

  8. Weaving order from uncertainty: Design, Analysis, and Applications of Transport-based Generative Models

    … that utilizes kernel methods to minimize Maximum Mean Discrepancy (MMD). Our method offers an efficient alternative to existing neural network approaches, achieving comparable performance with fewer parameters and reduced training time. We apply the theoretical results from the first part …

    washington Repository record for Weaving order from uncertainty: Design, Analysis, and Applications of Transport-based Generative Models (opens in a new tab)

  9. Methodological advancements for synthetic data generation

    … the Structures Laplacian Multi-distributional Maximum Mean Discrepancy (SLM MMD), a novel tabular fidelity metric. SLM MMD can be used for hyperparameter tuning, has attractive theoretical properties, and is less computationally expensive than existing methods. Together these three …

    bu Repository record for Methodological advancements for synthetic data generation (opens in a new tab)

  10. Statistically calibrating comparison metrics of star formation simulations.

    … spectral correlation function (SCF), Cramer, and maximum mean discrepancy (MMD). We ran a suite of carefully chosen simulations, and performed a simulation-to-simulation comparison to test the framework in a controlled environment. We have tested the effects of four physical parameters, Mach …

    ubc Repository record for Statistically calibrating comparison metrics of star formation simulations. (opens in a new tab) Download PDF for Statistically calibrating comparison metrics of star formation simulations. (opens in a new tab)

  11. Deep kernel mean embeddings for generative modeling and feedforward style transfer

    … is the combined use of neural networks and maximum mean discrepancy as a versatile similarity function. Additionally, we describe an improvement to state-of-the-art style transfer that allows faster computations while maintaining generality of the generating process. The proposed objective …

    ubc Repository record for Deep kernel mean embeddings for generative modeling and feedforward style transfer (opens in a new tab) Download PDF for Deep kernel mean embeddings for generative modeling and feedforward style transfer (opens in a new tab)

  12. A Robust Nonparametric Framework for Repeated Spatial Clustering: Insights from Spatial Omics

    … framework, we use a nonparametric test based on Maximum Mean Discrepancy (MMD) and block permutation to assess whether the distributions of multivariate attributes within the clusters are similar. This approach allows for re-partitioning of the clusters if necessary. To evaluate the proposed …

    mcmaster Repository record for A Robust Nonparametric Framework for Repeated Spatial Clustering: Insights from Spatial Omics (opens in a new tab)

  13. Robust and fair learning: From correlation matrices to fair learning models

    … generalised linear model framework that uses the maximum mean discrepancy as the fairness measure. Our choice of fairness measure can capture more complex differences between distributions from different sensitive groups. The model can be applied to datasets with different outcome types so is …

    city-london Download PDF for Robust and fair learning: From correlation matrices to fair learning models (opens in a new tab)

  14. Some statistical contributions to physics-informed neural networks and energy-based models

    … these short-run MCMC methods and kernel-based maximum mean discrepancy (MMD) estimations, specifically highlighting that the kernel used corresponds to the neural tangent kernel. The proposed framework is validated through practical examples, including applications to high-dimensional Gaussian …

    bu Repository record for Some statistical contributions to physics-informed neural networks and energy-based models (opens in a new tab)

  15. Assessing the Robustness of HAR Deep Learning Models against Variability

    … of how each affects DL HAR models' performance. Maximum Mean Discrepancy (MMD) was used to quantify shifts in data distribution due to each isolated variability and drew a relationship between the drop in performance and the change in data distribution. The REALDISP dataset was used to perform a …

    concordia Download PDF for Assessing the Robustness of HAR Deep Learning Models against Variability (opens in a new tab)

  16. Statistical inference for high-dimensional data

    … include the well-known energy distance and maximum mean discrepancy with Gaussian and Laplacian kernels, and the critical values are obtained via permutations. We show that all these tests are inconsistent when the two high dimensional distributions correspond to the same marginal …

    uiuc Repository record for Statistical inference for high-dimensional data (opens in a new tab)

  17. Manifold-Aware Regularization for Self-Supervised Representation Learning

    … and statistical dependency measures, such as maximum mean discrepancy (MMD) and Hilbert–Schmidt independence criterion (HSIC). Our contributions span both theory and practice. Theoretically, we unify contrastive and non-contrastive SSL objectives under a manifold-aware regularization …

    waterloo Repository record for Manifold-Aware Regularization for Self-Supervised Representation Learning (opens in a new tab)

  18. Distance-based Empirical Bayes: a nonparametric B-spline approach

    … marginal distribution using either the Maximum Mean Discrepancy or the standard 1-Wasserstein distance. Likelihood-aware knot placement allows the same construction to be used across Gaussian, Poisson, Binomial, Gamma, and Weibull observation models without reducing all problems to a …

    temple Repository record for Distance-based Empirical Bayes: a nonparametric B-spline approach (opens in a new tab)

  19. Generating Realistic Labeled Graphs with Conditional Generative Models

    … PROTEINS, ENZYMES, and MUTAG. Assessment using Maximum Mean Discrepancy (MMD) across degree, clustering, and spectral distributions demonstrates that our framework captures both local and global topology more effectively than existing baselines. Furthermore, high scores in uniqueness and novelty …

    brock Repository record for Generating Realistic Labeled Graphs with Conditional Generative Models (opens in a new tab)

Page 1 of 3