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 20 for “"high dimensional probability"”.
-
Problems in high-dimensional probability
In Chapter 1, we focus on lower estimates for suprema of stochastic processes. We investigate the role of super-gaussianity and the Sudakov Minoration Principle for this concept. We derive an extension of celebrated Majorizing Measures Theorem (MMT) for Gaussian mixtures which in turn yields a …
-
Fast and private distributed learning for low rank models
… from non-convex analysis, linear algebra, and high-dimensional probability, analyzed their convergence. The results of this research have been published in 3 peer-reviewed publications.
-
Deep Generative Models and Biological Applications
<p>High-dimensional probability distributions are important objects in a wide variety of applications. </p><p>Generative models provide an excellent manipulation method for training from rich available unlabeled data set and sampling new data points from underlying high-dimensional probability …
-
Spectral Independence A New Tool to Analyze Markov Chains
… of Markov chainMonte Carlo algorithms in high-dimensional probability and statistics. We rigorously prove rapid mixing of practically usefully Markov chains for sampling from important classes of probability distributions arising in computer science, statistical physics, and pure …
-
Particle Markov chain Monte Carlo
… emerged as the two main tools to sample from high-dimensional probability distributions. Although asymptotic convergence of MCMC algorithms is ensured under weak assumptions, the performance of these latters is unreliable when the proposal distributions used to explore the space are poorly …
-
SEQUENTIAL MONTE CARLO AND DIFFUSION-BASED SAMPLING METHODS: ALGORITHMS, LOSSES, AND EMPIRICAL ANALYSIS
… investigates methods for sampling from complex, high-dimensional probability distributions, with the goal of understanding their performance, limitations, and underlying mechanisms. Both established techniques and recent diffusion-based samplers are studied. Numerical experiments show that …
-
Distribution Optimal Importance Weights For Efficient Uncertainty Propagation Through Model Chains
… importance weights to achieve a change of probability measure. The objective of the thesis is to estimate statistics from a target distribution - distribution of interest using random samples generated from a different proposal distribution - cheap/available distribution. The approach taken …
-
LOCALIZATION IN SAMPLING: THEORY AND APPLICATIONS
Sampling from high-dimensional probability distributions is hindered by the curse of dimensionality (CoD), leading to prohibitive complexity. This thesis develops and analyzes the localization method, which exploits sparse dependency structures to design samplers whose complexity grows only mildly …
-
Generative compressed sensing with Fourier measurements
… subspace counting arguments and ideas central to high-dimensional probability. Furthermore, we propose a regularization strategy for training GNNs to have favourable coherence with the measurement operator. We provide compelling numerical simulations that support this regularized training …
-
An Uncoupling-Coupling Method for Markov Chain Monte Carlo Simulations with an Application to Biomolecules
… (UC), a new approach for drawing samples from high-dimensional probability distributions based on the Markov Chain Monte Carlo (MCMC) methodology. MCMC simulations are often hampered by strong metastabilities. This means that the sample path of a Monte Carlo Markov chain jumps rarely if at all …
-
Large-scale Phenomena in Geometry, Probability, and Combinatorics
… large-scale phenomena in the areas of geometry, probability, and combinatorics. The contributions of this thesis are as follows: In chapter 2 we study the $\ell_0$ isoperimetric coefficient for measurable sets in $\R^n$. Firstly we show that the $\ell_0$ isoperimetric coefficient of an …
-
Belief Propagation with Deep Unfolding for High-dimensional Inference in Communication Systems
<p>High-dimensional probability distributions are important objects in a wide variety of applications for example, most prediction and inference applications focus on computing the posterior marginal of a subset of variables conditioned on observations of another subset of variables. In practice, …
-
Non-reversible parallel tempering on optimized paths
… chain Monte Carlo schemes used to sample complex high-dimensional probability distributions. They rely on a collection of N interacting auxiliary chains targeting tempered versions of the target distribution to improve the exploration of the state-space. We provide here a new perspective on these …
-
A non-asymptotic study of low-rank estimation of smooth kernels on graphs
… dimension of the kernel. As is typical in high-dimensional statistics, we are able to design a suitable estimator based on a small number of samples only when the target kernel belongs to a subset of restricted complexity. In our study, we restrict the complexity by considering scenarios …
-
Multi-state Bayesian Process Control
… is represented by a belief vector within a high dimensional probability simplex. It is well known that obtaining structural results for such high-dimensional POMDP is challenging. Firstly, we prove that for an infinite-horizon process subject to multiple competing assignable causes, a …
-
High-dimensional Limit of SGD for Adaptive Stepsize Algorithms and Diagonal Linear Networks
… descent (SGD) at finite stepsizes through high-dimensional limits, where algorithmic randomness gives rise to deterministic evolution laws for key quantities of interest. We begin by studying adaptive stepsize methods in high-dimensional optimization problems, which we refer to as the high …
-
Generative modelling and adversarial learning
… machine learning is to represent and manipulate high-dimensional probability distributions of real-world data, such as natural images. Generative adversarial networks (GAN), which are based on the adversarial learning paradigm, are one of the main types of methods for deriving generative models …
-
Combining Reasoning and Learning for Multi-stage Inference in Computational Sustainability and Scientific Discovery
… NP, multi-stage inference problems are often highly intractable. In this thesis, I introduce a novel computational framework, based on embeddings, to tackle multi-stage inference problems. Our embedding technique approximates the intractable sub-problems of a multi-stage inference problem …
-
Superconvergence and accuracy enhancement of discontinuous Galerkin solutions for Vlasov-Maxwell equations and Numerical Analysis of a Hybrid Method for Radiation Transport
… stability properties. However, to resolve the high dimensional probability distribution function, the computational cost is the main bottleneck even for modern-day supercomputers. The first part of this thesis studies the applicability of a post-processing technique to the DG solution to …
-
Uncertainty quantification in high-dimensional models and post-selection procedures
… addressing uncertainty quantification in high-dimensional structured ensembles, high-dimensional false discovery control, and valid confidence intervals in post-selection inference. In the first project, we introduce the concept of structured high-dimensional probability simplexes …