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 23 for “"Fine-Tuning Strategies"”.
-
Efficient Fine-Tuning Strategies for Federated Learning: Optimizing Model Performance Across Distributed Networks
… strate its efficiency and combine it with full fine-tuning to produce stronger performance. Then, we introduce an adapted zeroth-order method capable of bringing a model to convergence with a mini- mal per-round compute budget while reducing the memory burden for clients during training down to …
-
A comparative analysis of large language models on medical datasets : performance evaluation and fine-tuning strategies
… regarding domain-specific benchmarks and fine-tuning with specialized datasets. This study addresses this gap by investigating the role of KD in enhancing LLM performance within the medical domain, where accuracy and contextual understanding are crucial. A dual-track approach is adopted: …
-
Self-Supervised Vision Transformers for Steel Surface Defect Detection - An Empirical Investigation of Fine-Tuning Strategies and Data Efficiency
… methods under different conditions, including fine-tuning vs linear probing, ImageNet initialization vs training from scratch and varying amounts of labeled data. A comprehensive experimental setup is used to assess both overall performance and label efficiency and results are compared to a …
-
Region-of-Interest Detection for the Neonatal Intensive Care Unit
… from adult data under different pretraining and fine-tuning strategies. Lastly, this thesis demonstrates that it is possible to estimate full pose from neonates even when the patient is fully occluded, using a combination of pressure and depth imagery.
-
EFFORTS TOWARD TRUSTWORTHY MACHINE LEARNING: MITIGATING OVERCONFIDENCE, HALLUCINATION, AND MODALITY BIAS
… challenges through post-hoc methods and targeted fine-tuning strategies. First, we mitigate overconfidence in discriminative models by introducing a neighborhood agreement metric that leverages foundation model representations, offering more reliable confidence estimates than softmax scores. We …
-
Analyzing Icelandic conversation using State-of-the-Art ASR models
This thesis investigates the performance and fine-tuning of transformer-based Automatic Speech Recognition (ASR) systems applied to Icelandic conversational speech, with a primary focus on OpenAI’s Whisper model and a secondary fo- cus on Meta’s Wav2Vec 2.0. The study uses the Spjallrómur dataset, …
-
Harmonizing Divergence in Computational Discourse Analysis
… IDRR. The thesis further explores prompting strategies using recent large language models and shows that fine-tuning strategies still perform better in this task. Together, these contributions advance the goal of harmonizing divergence in computational discourse analysis, offering more …
-
Exploration in Personalized Federated Learning under Heterogeneous Multi-modal Environments
… How can I design more effective personalization strategies to mitigate the adverse effects of non-IID data distributions? (2) Personalization for Missing Modality in Multimodal FL: How can I address the problem of missing modalities in multimodal federated settings without relying on auxiliary …
-
Toward Transformer-based Large Energy Models for Smart Energy Management
… architectures, utilizing both zero-shot and fine-tuning strategies, with data from 1,014 buildings. The results, evaluated across three prediction horizons (24, 72, and 168 hours), confirm that GEMs significantly outperform Transformer-based in-situ (i.e., building-specific) models. …
-
On Bed Posture Recognition Using Deep Learning With Pressure Sensors
… approaches to get over these restrictions, then fine-tuned it using the following deep learning architectures for bed posture recognition: ResNet-50, EfficientNet-B0, MobileNetV2, EfficientNet-B4, and ResNet-101. Our approach involves pressure sensor-based bed posture detection entails …
-
Sequence-to-Sequence Automatic Sentence Simplification: Development & Evaluation
… the research investigates the potential of fine-tuning pre-trained Large Language Models (LLMs) for simplifying complex sentences. It assesses the integration of LLMs into Seq2Seq models and their adaptation to the simplification task, aiming to leverage the benefits of pre-trained language …
-
Prior knowledge-driven model pre-training and fine-tuning for ophthalmic image analysis
… lesions, and the inefficiency of conventional fine-tuning strategies for large pre-trained models (LPMs). This thesis addresses these challenges by developing and evaluating novel pre-training and fine-tuning frameworks tailored for ophthalmic imaging. We first establish a strong and …
-
Efficient Multimodal Foundation Model Tuning For Hallucination Mitigation
… from fixed learning (static, offline fine-tuning) to adaptive, feedback-driven lifelong learning. By incorporating test-time adaptation, it seeks to enable multimodal foundation models to evolve through interaction with real data. In this process, an evaluator assesses the overall …
-
Privacy-Preserving AI for Mental Wellbeing Monitoring: From Passive Sensing to Online Engagement, through a User-Centred Lens
… retrieval-augmented<br/>generation, and fine-tuning strategies across model scales, and showing that compact fine-tuned<br/>models can achieve strong accuracy while remaining efficient enough to run directly on personal<br/>devices. This makes it possible to provide responsive, real-time …
-
Uncertainty-Aware and Data-Efficient Fine-Tuning and Application of Foundation Models
… (ELREA), designed for date-efficient multi-task fine-tuning, which clusters training instructions based on gradient directions and applies task-specific Low-Rank Adaptation (LoRA) experts through ensemble techniques. ELREA mitigates task interference, promoting better generalization and parameter …
-
A Multifaceted Analysis of Model Generalization in Challenging Environments
… architectures, training distributions, and fine-tuning strategies in Chapter 3. Beyond overall accuracy, the analysis examines robustness with respect to specific visual factors, out-of-distribution conditions, and the interaction between vision and language encoders. It further incorporates …
-
Anthrogames
… social-media datasets and optimized via specific fine-tuning strategies to ensure temporal robustness and cross-lingual generalization. Fifth, and independently of the safeguarding layer, the thesis implements and evaluates the continuous pretraining of a mid-scale multilingual large language …
-
Enhancing language models for telecom networks using retrieval-augmented generation: A pipeline for telecom-specific QA system
… as the interpretation of technical standards defined by organizations such as Third Generation Partnership Project (3GPP) standards, industry-specific jargon, and the need for precise technical answers necessitate specialized models and effective telecom question-answer systems. In this thesis, …
-
AI for Medical Imaging: Foundation Models, 2D-to-3D Reconstruction, and Clinical Applications
… problem.</p><p>This dissertation explores strategies for developing and applying medical image segmentation algorithms, with a focus on adapting vision foundation models to the medical domain and integrating segmentation into downstream clinical and technical applications.</p><p>In Chapter …
-
Robustness under distribution shifts in computer vision
… two distribution shifts and propose mitigation strategies to improve their robustness to the shifts. Prior Shifts: improving recall without sacrificing precision. Prior shifts can arise when i) the training class distribution does not match the true (test) class distribution due to the inherent …
Page 1 of 2