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 15 of 15 for “"Shortcut learning"”.

  1. From Shortcut Learning to Explainable Prediction: A Framework for Complex Pulmonary Disease Detection

    … of co-occurring diseases. Traditional deep learning approaches, which often focus on single-disease classification, fail to account for the complexities of multi-pathology presentations and raise concerns about bias and interpretability. This thesis addresses these limitations by advancing …

    whiterose Download PDF for From Shortcut Learning to Explainable Prediction: A Framework for Complex Pulmonary Disease Detection (opens in a new tab)

  2. Preregistration and Registration as a New Method for Transparency of External Validation in Artificial Intelligence and Machine Learning Applications to Address Overfitting, Underspecification, and Shortcut Learning

    … allows for a clear-cut distinction between the learning approaches used in the two cultures in all cases. Indeed, for decades, the two cultures have evolved within partly independent sociological niches. AI/ML can be seen as involving searching a very large space of possible hypotheses to …

    south-carolina Repository record for Preregistration and Registration as a New Method for Transparency of External Validation in Artificial Intelligence and Machine Learning Applications to Address Overfitting, Underspecification, and Shortcut Learning (opens in a new tab)

  3. Visual recognition of unseen objects and its challenges

    Conventionally, deep learning models learn to classify images of objects by training under the classification task on given training sets whose image samples are annotated with defined sets of target class labels. Conventional models can only classify an input image as an object class seen during …

    freiburg-diss Repository record for Visual recognition of unseen objects and its challenges (opens in a new tab)

  4. Automated Detection of Learning Pitfalls in Tabular Security Classifiers via Reasoning LLMs

    … yet they are frequently undermined by systemic learning flaws. Vulnerabilities such as shortcut learning and out-of-distribution (OOD) susceptibility can artificially inflate perceived performance, providing the illusion of a robust classifier while leaving real-world deployments vulnerable. …

    vt Repository record for Automated Detection of Learning Pitfalls in Tabular Security Classifiers via Reasoning LLMs (opens in a new tab)

  5. REPRESENTATION LEARNING FOR VISUAL TASKS: A STUDY OF ATTENTION AND INFORMATION SELECTION

    … methods for improving visual representation learning by optimizing attention mechanisms and information selection strategies within deep learning models. Standard approaches often process images independently and compress them into single global descriptors, limiting performance on tasks …

    temple Repository record for REPRESENTATION LEARNING FOR VISUAL TASKS: A STUDY OF ATTENTION AND INFORMATION SELECTION (opens in a new tab)

  6. Building Multimodal AI Systems for Perceiving and Reasoning in 3D Space

    … spatial competence from benchmark-specific shortcut learning. These three requirements---evidence-preserving representation, reasoning-explicit architecture, and evidence-sensitive evaluation---form the organizing framework of this work. The first part develops efficient representations for …

    washington Repository record for Building Multimodal AI Systems for Perceiving and Reasoning in 3D Space (opens in a new tab)

  7. Building robust and adaptive foundation models using structured learning signals

    Machine learning models are increasingly adapted in settings where supervision is incomplete, biased, or indirect, and where data distributions, tasks, and behavioral requirements evolve. In such regimes, optimizing the task objective alone often leads to spurious shortcut learning, catastrophic …

    ubc Repository record for Building robust and adaptive foundation models using structured learning signals (opens in a new tab) Download PDF for Building robust and adaptive foundation models using structured learning signals (opens in a new tab)

  8. Integrating Expertise: Domain-Driven AI Approaches for Cardiovascular Disease

    … calibration technique to detect and correct for shortcut learning caused by hidden data acquisition artifacts in medical datasets. Second, we propose domain-knowledge AI models that embed cardiovascular expertise into the models. For ventricular arrhythmia risk prediction, we incorporate QRS peak …

    toronto-retro Repository record for Integrating Expertise: Domain-Driven AI Approaches for Cardiovascular Disease (opens in a new tab)

  9. Towards High Performing and Reliable Deep Convolutional Neural Network Models for Typically Limited Medical Imaging Datasets

    … reducing costs and improving efficiencies. Deep learning is a subset of AI that is becoming transformative for healthcare. Deep learning offers fast and accurate data analysis. Deep learning is based on the concept of artificial neural networks to solve complex problems. In this dissertation, we …

    usf Repository record for Towards High Performing and Reliable Deep Convolutional Neural Network Models for Typically Limited Medical Imaging Datasets (opens in a new tab)

  10. Explainable AI for Network Intrusion Detection in Modern In-Vehicle Networks - An explainability pipeline for deep learning based intrusion detection systems for in-vehicle networks

    Deep learning–based network intrusion detection systems (NIDS) have been widely adopted for detecting attacks in Controller Area Network (CAN) traffic due to their superior performance over traditional approaches. However, their black-box nature makes the underlying decision-making process …

    chalmers Repository record for Explainable AI for Network Intrusion Detection in Modern In-Vehicle Networks - An explainability pipeline for deep learning based intrusion detection systems for in-vehicle networks (opens in a new tab)

  11. Multimodal deep learning methods for image-text verification in misinformation detection

    … contributing to the development of novel deep learning methods, the creation of dedicated datasets, and evaluation under realistic conditions. Despite progress in the field, significant challenges remain, such as unimodal biases, reliance on superficial shortcut patterns, the retrieval of …

    greece Repository record for Multimodal deep learning methods for image-text verification in misinformation detection (opens in a new tab)

  12. Evolving Language Patterns on the Web: Community Influence, Model Adaptation, and Bias Mitigation

    … we highlight the importance of mitigating shortcut learning in both pre-training and fine-tuning stages. Third, we investigate how emojis, originally designed for neutral or positive expression, are repurposed for offensive communication. We develop a multi-step LLM-based pipeline to …

    maryland Repository record for Evolving Language Patterns on the Web: Community Influence, Model Adaptation, and Bias Mitigation (opens in a new tab)

  13. Toward Causal Generative Modeling: From Representation Learning to Controllable Generation

    … observation and intervention. While modern deep learning has excelled at identifying statistical patterns, current generative models often struggle to capture the underlying structural causal mechanisms of the data-generating process, leaving them vulnerable to shortcut learning and spurious …

    arkansas Repository record for Toward Causal Generative Modeling: From Representation Learning to Controllable Generation (opens in a new tab)

  14. Explainable AI for Automatic Document Classification in Regulated Finance

    … against external documents, and potential shortcut learning effects. Since both the training and evaluation data were generated using the same controlled template-based document generation methodology, the results should be interpreted within the context of this experimental setting. …

    chalmers Repository record for Explainable AI for Automatic Document Classification in Regulated Finance (opens in a new tab)

  15. Domain Generalisation in Reinforcement Learning

    Deep reinforcement learning (RL) aims to learn a general policy for the agent to act in an environment. The policy learns state representations from observations using deep neural networks. As such, deep RL inherently suffers from their defects. For instance, neural networks suffer from shortcut …

    adelaide Repository record for Domain Generalisation in Reinforcement Learning (opens in a new tab)