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 26 for “"semantic embeddings"”.

  1. Biomedical Semantic Embeddings: Using Hybrid Sentences to Construct Biomedical Word Embeddings and its Applications

    Word embeddings is a useful method that has shown enormous success in various NLP tasks, not only in open domain but also in biomedical domain. The biomedical domain provides various domain specific resources and tools that can be exploited to improve performance of these word embeddings. However, …

    unt Repository record for Biomedical Semantic Embeddings: Using Hybrid Sentences to Construct Biomedical Word Embeddings and its Applications (opens in a new tab)

  2. A TWO-STAGE APPROACH TOWARDS GENERALIZED ZERO-SHOT LEARNING

    … information, such as textual descriptions or semantic embeddings, about those classes. In the inference phase, samples consist of classes that could be either seen or unseen during training. This study introduces a two-stage approach in handling the GZSL setting. In the first stage, we …

    nus Repository record for A TWO-STAGE APPROACH TOWARDS GENERALIZED ZERO-SHOT LEARNING (opens in a new tab) Download file for A TWO-STAGE APPROACH TOWARDS GENERALIZED ZERO-SHOT LEARNING (opens in a new tab)

  3. Using Data Analysis and Machine Learning for Studying and Predicting Depression in Users on Social Media

    … generate a more robust and reliable system where semantic embeddings trained from advanced neural networks were adopted under the tweet-level, whilst for the user-level, an approach using 12 significant features was operated by extensive feature engineering. Further, SVM with Word2Vec and TFIDF …

    carleton Repository record for Using Data Analysis and Machine Learning for Studying and Predicting Depression in Users on Social Media (opens in a new tab)

  4. Semantic Message Representations for Behavioral Prediction in a Diabetes Prevention System

    … This work replaces those identifiers with semantic embeddings from two BERT-based models, general-domain BERT and Sentence-BERT, evaluated across recurrent, convolutional, and attention-based architectures. Prediction accuracy remains modest across all conditions, and richer models do not …

    oregon Repository record for Semantic Message Representations for Behavioral Prediction in a Diabetes Prevention System (opens in a new tab)

  5. Conditional Neural Language Models for Multimodal Learning and Natural Language Understanding

    … introduce a LSTM encoder for learning visual-semantic embeddings for ranking the relevance of text to images in a joint embedding space. Next we introduce three log-bilinear models for generating image descriptions that integrate both additive and multiplicative interactions. Beyond image …

    toronto-retro Repository record for Conditional Neural Language Models for Multimodal Learning and Natural Language Understanding (opens in a new tab)

  6. Training Efficiency and Robustness in Deep Learning

    … models. In the context of learning visual-semantic embeddings, we find that prioritizing learning on more informative training data increases convergence speed and improves generalization performance on test data. We formalize a simple trick called hard negative mining as a modification to …

    toronto-retro Repository record for Training Efficiency and Robustness in Deep Learning (opens in a new tab)

  7. Data representations in high-dimensional and complex data: predictive modeling in healthcare and social media

    … line of work by incorporating sentence-level semantic embeddings into graph-based user modeling. The final two studies further extend this direction. MELT combines topic features and transformer embeddings with time-weighted weak labels, achieving the strongest macro F1 among prior graph-based …

    temple Repository record for Data representations in high-dimensional and complex data: predictive modeling in healthcare and social media (opens in a new tab)

  8. Enhancing the effectiveness of information retrieval techniques for known-item retrieval from lifelogs

    … presented use pre-trained textual and visual-semantic embeddings for the lifelog retrieval problem and enhance retrieval performance with a high-level semantic mapping of user information. In summary, this thesis primarily contributes to the field of building an effective and efficient lifelog …

    dcu Download PDF for Enhancing the effectiveness of information retrieval techniques for known-item retrieval from lifelogs (opens in a new tab)

  9. Utilizing large language models and attention networks as AI copilots for enhanced decision-making in healthcare and defense

    … expansion, dense retrieval using domain-adapted semantic embeddings, Reciprocal Rank Fusion (RRF), and cross-encoder re-ranking. This architecture balances the advanced reasoning capabilities of modern LLMs with the computational efficiency and strict accuracy required in zero-fail environments. …

    umkc Repository record for Utilizing large language models and attention networks as AI copilots for enhanced decision-making in healthcare and defense (opens in a new tab)

  10. Context-Aware Structural Generalization for API Usage Pattern Mining

    … into the structural mining pipeline to assess semantic similarity. By dynamically adjusting the structural merge threshold based on CodeBERT semantic distance, CASGard combines the precision of AST-based generalization with the contextual information provided by neural semantic embeddings. …

    calgary Repository record for Context-Aware Structural Generalization for API Usage Pattern Mining (opens in a new tab)

  11. The Neural Bases of Object Memory: Neuromodulation and Hemodynamic Modeling in Mild Cognitive Impairment

    … is shaped by a variety of factors, including the semantic features of those stimuli and the fidelity with which the features are represented in the brain. This dissertation investigates the role of stimulus representations in object memory across the adult lifespan, combining behavioral, …

    duke Repository record for The Neural Bases of Object Memory: Neuromodulation and Hemodynamic Modeling in Mild Cognitive Impairment (opens in a new tab)

  12. Computer-Verified Foundations of Metaphysics and an Ontology of Natural Numbers in Isabelle/HOL

    We utilize and extend the method of shallow semantic embeddings (SSEs) in classical higher-order logic (HOL) to construct a custom theorem proving environment for abstract objects theory (AOT) on the basis of Isabelle/HOL. SSEs are a means for universal logical reasoning by translating a target …

    fu-berlin Repository record for Computer-Verified Foundations of Metaphysics and an Ontology of Natural Numbers in Isabelle/HOL (opens in a new tab)

  13. A survey of diffusion models and their applicabillity to digital staining of histological specimens

    … et al., 2021) model, renowned for its semantic richness, we adapt its capabilities to meet the high-resolution demands typical of histology. Preliminary results from our training experiments demonstrate this approach’s potential. Furthermore, we introduce a novel methodology named …

    bu Repository record for A survey of diffusion models and their applicabillity to digital staining of histological specimens (opens in a new tab)

  14. Open doors, closed debates : the evolution of responsiveness in 1980s Singapore

    … (NLP) tools, and then analyzed through semantic embeddings generated by the Ada-002 model. The alignment between public letters and parliamentary records, quantified via calculating Euclidean distance between embeddings, provides a measure of policy responsiveness. Other pieces of …

    ubc Repository record for Open doors, closed debates : the evolution of responsiveness in 1980s Singapore (opens in a new tab) Download PDF for Open doors, closed debates : the evolution of responsiveness in 1980s Singapore (opens in a new tab)

  15. AI-Powered Recommender System for Clinical Trial Protocol Design - A Tool for Medical Practitioners

    … end points form a reviewed hierarchy using semantic embeddings, hierarchical clustering, terminology-assisted standardization, and LLM-assisted review. Protocol and end point information is standardized against NCIt, CDISC, and LOINC for a two-stage recommendation pipeline: Stage 1 predicts …

    chalmers Repository record for AI-Powered Recommender System for Clinical Trial Protocol Design - A Tool for Medical Practitioners (opens in a new tab)

  16. GENERATIVE MODELS IN NATURAL LANGUAGE PROCESSING AND COMPUTER VISION

    … generation model with extended phonetic and semantic embeddings (Phonetic CNN_subword embeddings). Extensive quantitative experiments using BLEU scores and Hamming distance show notable enhancements over strong baselines. Additionally, a comprehensive human evaluation confirms that the poems …

    siu-theses Repository record for GENERATIVE MODELS IN NATURAL LANGUAGE PROCESSING AND COMPUTER VISION (opens in a new tab)

  17. Zombies Reading Segmented Graphene Articles On The Arxiv

    I present results obtained in four very distinct areas. In Chapter 1, I investigate vector based embedding models for text as applied to a corpus of scientific articles from the ar[chi]iv . I report on the utility of the learned word representations, and experiment with several techniques for …

    cornell Repository record for Zombies Reading Segmented Graphene Articles On The Arxiv (opens in a new tab)

  18. Context-aware Learning from Partial Observations

    … for ads and queries, where Deeply Supervised Semantic Match model is introduced for dealing with unobserved and tail queries click-through-rate prediction problem, while jointly learning the semantic embeddings of a query and an ad, as well as their corresponding click-through-rate. Finally, …

    temple Repository record for Context-aware Learning from Partial Observations (opens in a new tab)

  19. IMPROVING PERFORMANCE OF INTRUSION DETECTION SYSTEMS FOR SOFTWARE-DEFINED NETWORKS

    … a novel alert-based system that leverages semantic embeddings and prototypical networks to detect multi-stage attacks across different operational domains. By addressing domain shift challenges, CrossAlert ensures high performance and adaptability in diverse environments. The effectiveness …

    temple Repository record for IMPROVING PERFORMANCE OF INTRUSION DETECTION SYSTEMS FOR SOFTWARE-DEFINED NETWORKS (opens in a new tab)

  20. Multi-Perspective Feature Learning for Facial Expression Recognition in the Wild

    … of information, such as multiple levels of semantic patterns, facial geometry captured in facial landmarks, and multimodal representation of facial expressions, can be effectively extracted from the same facial image and integrated to enrich expression-discriminative feature representations. …

    south-carolina Repository record for Multi-Perspective Feature Learning for Facial Expression Recognition in the Wild (opens in a new tab)

Page 1 of 2