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 19 of 19 for “"Data Labelling"”.
-
Reducing the Cost of Test Data Labelling for Deep-Learning Systems: An Empirical Study
… comprehensive testing with accurately labeled datasets, which is a resource-intensive and costly process, especially for large-scale data. This thesis presents an empirical study aimed at reducing the cost of test data labeling for deep-learning systems through automated and semi-automated …
-
Efficient style-sensitive natural language processing: enhancing language model’s stylometric distinction and generation through resource-limited training and token probability manipulation
… plays an important role in content generation, data augmentation and misinforma-tion detection. Improving language models’ sensitivity to the text style often requires non-trivial effort on acquisition of large amount of labelled training data. In this paper, we propose a cost-effective data …
-
Activity and Gait Phase Recognition for Walking, Stair Ascent, and Stair Descent
… and was developed using inertial measurement data from 80 participants. It simultaneously detected activity and gait phase with 91.2% accuracy and 22.9% of perfectly classified strides. Results from this thesis support the use of FSRs for data labelling for machine learning-based recognition …
-
Diagnosis for Patient and GP: Dialogue-based Self-Diagnosis with Disease-Symptoms Graph and Referral Letter Classification
… of artificial intelligence (AI) and big data. The first strategy involves automated self-diagnosis tools for patients, which can help reduce the initial diagnostic burden on medi- cal systems. The second strategy targets General Practitioners (GPs), aiming to bolster their diagnostic …
-
Deep learning for industrial applications
… validated in the lab.</p> <p>We then identified data labelling to be a potential bottleneck in extending the solution to other applications, which led us to develop few-shot methods to address this. First, we introduced a method for interpolating batch normalization parameters across …
-
Novice-Centric Visualizations for Machine Learning
… system to help machine learning novices perform labelling tasks. There are many mature visualizations to help people understand the performance of current classifiers, including scatterplots, confusion matrices and ROC curves. However, most machine learning researchers are unaware of the …
-
Machine Learning for Hand Gesture Classification from Surface Electromyography Signals
… the current gesture based on incoming sEMG data. This thesis presents a set of methods, techniques and designs that improve upon evaluation of, and performance on, the classification problem as a whole. These are brought together to set a new baseline for the potential classification. …
-
Application of Artificial Neural Nets to a Selective Laser Sintering Additive Manufacturing Process For Fault Detection
… analysis of layer-by-layer surface profile data of Fe3Si powder. The data was collected in-situ during the metal additive manufacturing of a Heat Exchanger section, comprised of a series of conformal channels. Specifically, a shallow artificial neural net (ANN) was trained with …
-
3D Object Detection for Road Safety at Urban Intersections
… million injuries annually worldwide. Using crash data and traditional methods, it can take years and even decades to collect enough data to evaluate the safety of complex road infrastructure, such as intersections. Long data collection times necessitate a proactive approach using surrogate …
-
Supernova discovery in the machine learning era
… within transient astronomy, with key focus on data labelling, real/bogus and contextual classification. We highlight a conceptual framework for understanding the epistemic implications of automating discovery, and identify practices that preserve opportunities for understanding. We showcase …
-
Automated Theme Extraction from Homeless Shelter Case Notes: A Comparative Evaluation of Machine Learning and Large Language Model Methods
… for deriving population-level insights from data that are rich in contextual information but challenging to interpret at scale. Frontline staff often document client interactions and behaviours in these notes in an unstructured, narrative format. The considerable variability of case notes …
-
Pose-based computer vision detection of lameness in dairy cattle
… to find the best tool for video annotation, data labelling, and team projects. Among the available pose estimation model platforms, YOLO (You Only Look Once) was selected because of its balance of speed, accuracy, and its potential to do real time detection. In addition, training and …
-
BLE-based Indoor Localization and Contact Tracing Approaches
… environments. In this regard, first, an IoT dataset is constructed based on multiple thoroughly separated indoor environments to incorporate the effects of various interferences faced in different spaces. The constructed dataset is then used to develop a Reinforcement Learning (RL)-based …
-
Data-efficient Semantic Segmentation with Deep Learning
… deep learning continues to push the boundary of data science, solving tasks previously considered extremely difficult, if not impossible. Semantic segmentation, a pixel-wise dense classification task, is one of the computer vision tasks that deep learning models excel at. Accurate semantic …
-
ACOUSTIC EMISSION MONITORING OF THE POWDER BED FUSION PROCESS WITH MACHINE LEARNING APPROACH
… signals received during the printing and data collection steps are analyzed using an AE sensor under various process parameters. Several time and frequency-domain features were extracted from data during the mining process from the AE signals. K-means clustering is employed during …
-
Efficient Deep Learning Framework Using Convolutional Recurrent Neural Network for Elderly Care Centers for Fall Detection
… activities using Channel State Information (CSI) data, both with and without automatic data labeling. Experiments conducted without data labeling revealed that the CNN model, trained on pre-processed CSI data, achieved a validation accuracy of 85.33% and a testing accuracy of 84.52%, though it …
-
Inspecting the Machine Learning Pipeline: from Weak Supervision to Explainability
… starts with the collection and annotation of data or with a<br/>provided curated dataset that needs to be understood. Before any learning takes place, the first<br/>concern is data quality as this affects any subsequent tasks and inferences [67]. This has led to<br/>the development of tools …
-
As dry as a (burnt) bone – histological, physicochemical, and osteogenic reaction-related approaches to identifying the pre-burning condition of bone
… compared to its collagen content.</p> <p>Data analyses were carried out using a variety of methods. Histotaphonomy was assessed quantitatively and systematically by a novel data labelling application developed for this study in Python and Javascript programming languages. The …
-
Designing Software for Clinical Environments
… of these environments generates large volumes of data which is aggregated in clinical information systems and exposed to intensive care personnel using a variety of software tools which persist, organise and visualise the data to inform their decisions and allow them to adequately care for their …