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 187 for “"Data-Efficiency"”.
-
Improving Data Efficiency and Availability for Alignment
… thesis presents several projects which improve data efficiency and data availability for aligning AI systems using human feedback. First, it analyzes the exploration and exploitation tradeoff induced by introducing an explicit cost for observing rewards in a bandit setting. Second, it provides …
-
MEDICAL MACHINE INTELLIGENCE: DATA-EFFICIENCY AND KNOWLEDGE-AWARENESS
… challenging small targets, insufficient training data, high annotation cost, the lack of domain-specific knowledge, etc. These challenges cultivate the need for developing data-efficient and knowledge-aware deep learning techniques which can generalize to different medical tasks without requiring …
-
IMPROVING MODEL AND DATA EFFICIENCY FOR DEEP LEARNING
… considerable computation resources and expensive data collection costs. Various studies have strived to enhance efficiency in these domains. In terms of model efficiency, remarkable advancements have been made to accelerate the training and inference by methods such as quantization and pruning. …
-
Towards Data Efficiency and Robustness of Reinforcement Learning
… of RL, specifically in the areas of data efficiency and robustness. Part I of the dissertation delves into the data efficiency of reinforcement learning algorithms, with a primary focus on determining the optimal amount of data required for their implementation. In particular, we …
-
Probabilistic language models with model efficiency and data efficiency
… of applications, they typically demand vast datasets. In this study, we suggest an active learning method to label samples using an acquisition function with local sensitivity and learning difficulty. This method creates data replicas via subtle perturbations, prioritizing those data points …
-
On the Data Efficiency and Model Complexity of Visual Learning
… neural networks usually requires large-scale datasets annotated by humans, and the models typically have millions of parameters and consume a lot of computation resources. The issues of data efficiency and model complexity are commonly observed in many frameworks based on deep neural networks, …
-
Breaking the Bottlenecks of Scaling: Data Efficiency, Predictability, and Risks
… further scalability --- limited high-quality web data can bottleneck data scaling, model behavior at larger scales can become increasingly unpredictable, and more capable models can pose greater risks. This thesis aims to develop scalable methods to address these challenges. Each chapter addresses …
-
Data augmentation and data efficiency for low-resource language processing
Submission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2027-05-01
-
Data efficiency and domain robustness in speech-based biomarking of health
… A key challenge is domain robustness under data scarcity: medically labeled speech databases are small, and training–testing mismatches in speaker-level and environmental factors often degrade performance. This dissertation targets these issues in the context of classification tasks, …
-
Learning-based Autonomous Driving with Enhanced Data Efficiency and Policy Training
… with little to no human intervention, enhancing efficiency and safety. Self-driving cars, for instance, will affect our modes of transportation and way of life in the years to come. With rapid advances in hardware and software design, learning-based autonomous driving is becoming a viable and …
-
Leveraging Models to Improve Data Efficiency: Navigation, Reinforcement Learning, and Lie Group Convolutions
Consider a system which takes data as an input, processes the data with a model, and outputs a decision for a particular objective. We call the measure of the amount of data used to complete the objective with some performance metric as data efficiency. Across many domains, it is advantageous to …
-
Towards Generalized 3D Object Detection for Autonomous Driving: Enhancing Data Efficiency in Novel Domains
… supervised learning using large-scale annotated datasets, which are meticulously designed to reflect diverse real-world scenarios and object variations. Creating and annotating such datasets is notoriously labor-intensive and time-consuming, requiring significant efforts to ensure precision and …
-
CHALLENGES AND APPLICATIONS OF FINE-GRAINED TEMPORAL ACTION UNDERSTANDING: MODELING TEMPORAL GRANULARITY AND DATA EFFICIENCY
… remains prohibitively expensive, with most datasets requiring exhaustive labeling of lengthy untrimmed videos. These challenges are particularly pronounced in medical domains, where sub-second primitives define clinical outcomes while expert annotation remains scarce. In this dissertation, …
-
Detection of the uniqueness of a human voice: towards machine learning for improved data efficiency
The aim of this thesis is to characterise voice characteristics that can establish the identity of the person who is speaking, independent of the language used. The fundamental goal of the work is to understand how humans recognise a speaker. The voice parameters such as: speech rate, natural …
-
Self-Supervised Vision Transformers for Steel Surface Defect Detection - An Empirical Investigation of Fine-Tuning Strategies and Data Efficiency
… However, ob taining large amounts of labeled data is often costly and time-consuming, motivating the use of self-supervised learning (SSL) to leverage unlabeled data. This thesis in vestigates the effectiveness of SSL for defect classification using Vision Transformer based methods, with a …
-
ENHANCING DATA EFFICIENCY AND PHYSICAL CONSISTENCY OF GRAPH NEURAL NETWORKS FOR ASSESSING CUSTOMIZED AND 3D-PRINTING ENABLED CONCRETE PARTITION WALLS
This thesis explores the application of enhanced Graph Neural Networks (GNNs) as advanced surrogate evaluators in the architectural engineering domain, specifically for the development and analysis of customized concrete partition walls made with 3D-print-enabled approach. These structures are …
-
Efficient Learning with Subgoals and Gaussian Process
This thesis demonstrates how data efficiency in reinforcement learning can be improved through the use of subgoals and Gaussian process. Data efficiency is extremely important in a range of problems in which gathering additional data is expensive. This tends to be the case in most problems that …
-
Data- and compute-efficient visual recognition and generation
… of deep learning models escalates, achieving efficiency in both architecture construction and data utilization becomes paramount. This dissertation examines two fundamental categories of efficiency: model efficiency and data efficiency. 1. Model Efficiency: This facet of the study focuses on …
Page 1 of 10