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 147 for “"Noisy-OR"”.
-
Near-optimal sample complexity for noisy or 1-bit tensor completion
Tensor completion is the problem of recovering a low-rank tensor from a partial subset of its entries. Assuming a rank-r, order-d tensor in ℝ^{NxNx...N}, the best sampling complexity achieved is O(rN^{d/2}) which can be obtained by a tensor nuclear-norm minimization problem. This bound is …
-
Assessing the Quality of Synthetic Speech when using Enhanced Speech as Training Data
… fields, but their interaction remains under-explored. In particular, the effectiveness of using enhanced speech to train a speech synthesis model is still relatively unknown. This thesis investigates the effects of using enhanced speech to train a speech synthesis model on the overall quality of …
-
Deep Learning Based CT Image Reconstruction
… image created by conventional reconstruction algorithms will be noisy, or contain artifacts. The goal of our study is to reconstruct high-quality tomographic images from noisy or incomplete scan data, including low-dose, sparse-view, and limited-angle scenarios, by utilizing novel deep learning …
-
Intelligent Fusion of Evidence from Multiple Sources for Text Classification
… using current approaches is known to perform poorly when documents are noisy or when limited amounts of textual content is available. Yet, many users need access to such documents, which are found in large numbers in digital libraries and in the WWW. If documents are not classified, they …
-
Improved Scalability and Accuracy of Bayesian Network Structure Learning in the Score-and-Search Paradigm
A Bayesian network is a probabilistic graphical model that consists of a directed acyclic graph (DAG), where each node is a random variable and attached to each node is a conditional probability distribution (CPD). A Bayesian network (BN) can either be constructed by a domain expert or learned …
-
Modeling and evaluation of trusts in Multi-agent systems
… an approach to help agents construct a trust network automatically in a multi-agent system. After the construction of the trust network, we use the Bayesian Inference Propagation approach with Leaky Noisy-Or model to solve the trust graph.The experiments carried out by our simulation work …
-
The Little Diagram That Could: Geometric Properties and Statistical Applications of Persistence Diagrams in Topological Data Analysis
… Analysis (TDA) is a collection of techniques for data analysis that leverages topological invariants of spaces formed from data points. These methods excel at extracting useful information from noisy or sparse data, making them attractive to many mathematicians, statisticians, and scientists. …
-
Understanding how Language Models Learn Under Constrained or Noisy Settings
Transformer-based Language models (LMs) have become the foundation of modern Natural Language Processing (NLP), demonstrating strong performance across various tasks. Typically trained on massive datasets with curated objectives and clean supervision, these models are increasingly being deployed in …
-
Reconstructing pairwise comparisons matrices based on differential evolution: a Monte Carlo study
… challenge to generate a PC matrix from noisy or incomplete real-life input data. This study aims to investigate the reconstruction of pairwise comparisons matrices from not-so-inconsistent pairwise comparisons matrices by an optimization method based on di↵erential evolution. A …
-
Multiplicative factorization of multi-valued NIN-AND tree causal models
Bayesian networks (BNs) are compact representations of uncertain knowledge, which combine probability theory and graph theory. However, the size of each conditional probability table (CPT) in BNs still grows exponentially on the number of causes. Causal independence models (CIMs) are proposed to …
-
Enhancing Post Silicon Visibility Using Language Modelling Techniques
… validation phase is one of the most important, as it allows engineers to test the behavior of a device in a real world setting. However, the issue of noisy or incomplete data is a frequent issue when attempting to debug an SoC design during this step. This thesis examines the utility of …
-
An Empirical Study of NIN-AND Tree Elicitation
Constructing a Bayesian Network requires the conditional probabilities table (CPT) to be acquired, one for each variable or node in the network. When data mining is not available, CPTs must be acquired from the domain experts. The complexity of the direct elicitation is exponential on the number of …
-
MULTI-SOURCED INFORMATION TRUSTWORTHINESS ANALYSIS: APPLICATIONS AND THEORY
… data entries, even describing the same objects or events, can come from a variety of sources. There are some sources that typically provide accurate information, but due to various reasons such as recording errors, device malfunction, background noise and intent to manipulate the data, some …
-
Efficient sparse Bayesian machine learning in nonlinear aeroelasticity using wind tunnel data
… capability of mechanistic and neural network models, trained using experimental (wind tunnel) data. Calibration of overly complex models with sparse, noisy, or incomplete observations can lead to overfitting, compromising their predictive performance. To this end, Bayesian sparse learning …
-
Uncertainty and sensitivity analysis in support vector machines: Robust optimization and uncertain programming approaches.
Noisy or uncertain data are common in machine learning and data mining applications. Noisy data can significantly affect the behavior of data mining and machine learning algorithms. Robust optimization and sensitivity analysis techniques are applied to the support vector machine (SVM) learning …
-
BAYESIAN TRANSFER LEARNING FOR PASSIVE SONAR TARGET CLASSIFICATION
… learning applications when datasets are too noisy or difficult to label. Furthermore, neural networks often lack explainability metrics to justify their predictions. In this work, we construct a framework for developing Bayesian classification models using transfer learning to detect ship …
-
Solving inverse problems with generative priors trained on corrupted data
This thesis explores machine learning-based methods for solving inverse problems in signal processing, focusing on applications in medical imaging and wireless communications. Specifically, I investigate the use of score-based generative models and diffusion models trained on corrupted data to …
-
Neural nets : a study of factors affecting the ability of a back propagation net to recognize alphabetic characters and produce specific symbols
… The net learns from the training pairs, and stores these relationships. The net provides a method to generalize an output based on the nearness to all of the trained inputs. This works well even in where the inputs are noisy or distorted. This is a study of a back propagation net. Neural nets …
-
Character recognition using string matching
<p>To handle noisy and distorted pattern is the use of similarity or distance measures. A similarity or distance measure can be defined between a representation of an unknown pattern and a representation of a prototype pattern. Recognition of the unknown pattern can be carried out on the basis of …
-
Solving Multiple Inference in Graphical Models
For inference problems in graphical models, much effort has been directed at algorithms for obtaining one single optimal prediction. In practice, the data is often noisy or incomplete, which makes one single optimal solution unreliable. To address this problem, multiple Inference is proposed to …
Page 1 of 8