End-to-end automated speech recognition using a character based small scale transformer architecture.

Get full text
Get full text

Fitness landscape analysis of feed-forward neural networks

Get full text

Deep Reinforcement Learning-Based Approaches for the MPPT Control of Standalone Solar PV Systems by Nwachukwu, Sampson

“…This is due to the introduction of robust maximum entropy in the SAC's learning process, which ensures that the system model and estimation errors are minimized, especially in complex control problems. …”
Get full text

Learning to speak and hear through multi-agent communication over a continuous acoustic channel by Eloff, Kevin

Get full text

A comparison of Gaussian mixture variants with application to automatic phoneme recognition by Brand, Rinus

Get full text

Measuring, refining and calibrating speaker and language information extracted from speech by Brummer, Niko

Get full text

An Artificial Intelligence Framework for Assessing the Financial Performance and Valuation of Construction Contractors in Egypt by Youssef, Omar Magdy

“…The first model provided mean absolute percentage error (MAPE) ranging between 20.62% to 72.8% for the tested companies. …”
Get full text
Get full text

Modelling of the deaerator system in Flownex by Banda, Richard Bobby

“…The study of the steady-state and dynamic behaviour of thermal power plants is of interest and significant benefit in different engineering fields ranging from research and design, to the assistance of operator training, plant optimization, fault finding and failure analysis. …”
Get full text

The role of academic libraries in implementing research data services: a case study of the University of KwaZulu-Natal Libraries by Madibi, Zizipho

“…The researchers who responded to the study showed interest in different trainings such as, training on data storage, development of DMPs and metadata creation. …”
Get full text

Application of CNN-gcForestCS to cassava leaf image classification by Carew, Liam

“…However, CNNs require complex hyperparameter tuning and can be computationally intensive to train. GcForestCS (multi-grained cascade forest with confidence screening) presents an alternative statistical learning method that can be trained using CPU, and requires less complex hyperparameter tuning than deep learning while producing competitive performance for lower-dimensionality datasets. …”
Get full text

Adaboost and its application using classification trees

Get full text

Factors influencing safe patient care provided by professional nurses in a private healthcare organisation in the Western Cape by Daniels, Ruth Carol

Get full text

Modelling of polymer seal strength: response surface methodology by Van Jaarsveld, Ivan

Get full text

Handwritten signature verification using hidden Markov models by Sindle, Colin

Get full text

First-year higher education students’ acquisition of digital content creation literacies in discipline-specific settings by Mayisela, Thabisa

“…The main contribution of this thesis is the finding that, when LAAs, which foster the acquisition of digital content-creation literacies are part of the curriculum, students can acquire the respective literacies through self-teaching, scaffolding from peers or senior students, trial and error, and engaging in practice within their disciplines, implying that there is no need for overt instruction as is the case of digital skills training isolated from the context of practice. …”
Get full text

Development of a human digital twin for human-centric dispatching for an assembly process by Kneissl, Anja

Get full text

Predictive Maintenance of Wind Turbines: Remaining Useful Life Estimation Using a BPSO-Optimized Attention-Based CNN-BiLSTM Approach by Hawas, Eman Khaled

“…The evaluation metrics used were the Mean Squared Error (MSE), Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), and the Coefficient of Determination (R²). …”
Get full text
Get full text

South African cranial variation : a combined metric-macromorphoscopic method for ancestry estimation

Get full text

Knowledge of South African anaesthesia registrars of the theory and practice of spinal anaesthesia for caesarean section by Ras, Willem

“…Results A total of 126/400 responses were received, with a response rate of 31% and margin of error 7.3%. Participants' questionnaire scores were not normally distributed. …”
Get full text