Federated Learning with Rebalanced Dataset

With the widespread application of artificial intelligence, centralized machine learning approaches, which require access to users' local data, have raised concerns about data privacy. In response, federated learning, an architecture that aggregates models trained locally with local data, has been proposed. This approach addresses the data...

Towards a low-cost air-written character recognition system: Designing an embedded machine learning system to recognise the first 10 letters of the Latin alphabet

This study introduces a novel system that leverages three photodiodes and ambient light to identify air-written characters on a resource-constrained device. Through experimentation, suitable methods of data preprocessing, machine learning and model compression were selected to recognize the first 10 characters of the Latin alphabet. The final...

Multi-level optimisation and global sensitivity analysis of the probabilistic damage stability method for single hold ships

The probabilistic damage stability method offers great design freedom when used as a base of design. However, due to the complexity of the calculation and amount of parameters that influence the attained index, much of this freedom is not being harnessed by designers. This research tries to give the designer more insight in where to look when...

Impact of Safe Return to Port regulatory framework on cruise ship concept design: A software tool to mitigate design risks in early stages of the process

The expansion of the cruise market in the last decades and the significant increase in the size of cruise ships, led to a revision of the safety standards for passenger vessels which resulted in the introduction of the so-called "Safe Return to Port" regulatory framework (SRtP). These regulations strongly impacted every aspect of the life of...