The COVID-19 pandemic shed light on the need for quick diagnosis tools in healthcare, leading to the development of several algorithmic models for disease detection. Though these models are relatively easy to build, their training requires a lot of data, storage, and resources, which may not be available for use by medical institutions or could be beyond the skillset of the people who most need these tools. This paper describes a data analysis and machine learning platform that takes advantage of high-performance computing infrastructure for medical diagnosis support applications. This platform is validated by re-training a previously published deep learning model (COVID-Net) on new data, where it is shown that the performance of the model is improved through large-scale hyperparameter optimisation that uncovered optimal training parameter combinations. The per-class accuracy of the model, especially for COVID-19 and pneumonia, is higher when using the tuned hyperparameters (healthy: 96.5%; pneumonia: 61.5%; COVID-19: 78.9%) as opposed to parameters chosen through traditional methods (healthy: 93.6%; pneumonia: 46.1%; COVID-19: 76.3%). Furthermore, training speed-up analysis shows a major decrease in training time as resources increase, from 207 min using 1 node to 54 min when distributed over 32 nodes, but highlights the presence of a cut-off point where the communication overhead begins to affect performance. The developed platform is intended to provide the medical field with a technical environment for developing novel portable artificial-intelligence-based tools for diagnosis support.

No abstract

Sound localization is the ability of humans to determine the source direction of sounds that they hear. Emulating this capability in virtual environments can have various societally relevant applications enabling more realistic virtual acoustics. We use a variety of artificial intelligence methods, such as machine learning via an Artificial Neural Network (ANN) model, to emulate human sound localization abilities. This paper addresses the particular challenge that the training and optimization of these models is very computationallyintensive when working with audio signal datasets. It describes the successful porting of our novel ANN model code for sound localization from limiting serial CPU-based systems to powerful, cutting-edge High-Performance Computing (HPC) resources to obtain significant speed-ups of the training and optimization process. Selected details of the code refactoring and HPC porting are described, such as adapting hyperparameter optimization algorithms to efficiently use the available HPC resources and replacing third-party libraries responsible for audio signal analysis and linear algebra. This study demonstrates that using innovative HPC systems at the Jülich Supercomputing Centre, equipped with high-tech Graphics Processing Unit (GPU) resources and based on the Modular Supercomputing Architecture, enables significant speed-ups and reduces the time-to-solution for sound localization from three days to three hours per ANN model.

To test generalization ability of a class of deep neural networks, we randomly generate a large number of different rule sets for 2-D cellular automata (CA), based on John Conway's Game of Life. Using these rules, we compute several trajectories for each CA instance. A deep convolutional encoder-decoder network with short and long range skip connections is trained on various generated CA trajectories to predict the next CA state given its previous states. Results show that the network is able to learn the rules of various, complex cellular automata and generalize to unseen configurations. To some extent, the network shows generalization to rule sets and neighborhood sizes that were not seen during the training at all.Preprint. Under review.