Mohsen Pourjam
Title of the Doctorate Thesis
Analysis of gut microbiome time-series data using Deep Neural Networks
Project description
The human gut microbiome is highly dynamic on different longitudinal timescales. This concerns short-term changes due to diet or medical interventions, and long-term changes, for instance, in childhood development. Thus, any given previous change and current composition could be used to delineate a trajectory concerning future compositions and as well as health conditions from children maturing into adulthood.
With the increased availability of time-series data of gut microbiomes from human studies, Time Series Classification (TSC) algorithms could give invaluable insights concerning health outcomes. Among TSC algorithms, Deep Neural Networks (DNNs) have rarely been considered even though DNNs are known to be a powerful technique successfully applied to a number of research questions in the last years. We expect DNNs will help to untangle the many correlations found in time-series data classification problems in order to understand gut microbiome dynamics.