This project is part of a Marie Curie Innovative Training Network called "Machine Learning Frontiers in Precision Medicine", in which 13 nodes across Europe interact while working on different aspects of the same problem. In particular, we aim to explore and develop methods for unsupervised learning on genetics and high-dimensional longitudinal data coming from different sources, such as DNA methylation, fMRI studies and even motion tracking of patients and disease-related experimental animals. While given the nature of our institution the original plan sticks to psychiatry-related data, the methods and algorithms to be developed are immediately applicable in other disease areas and data domains.
As the dimensionality of the data we use tends to be ultra-high, we are exploring pipelines involving state-of-the-art dimensionality reduction techniques before feeding the data to clustering algorithms. Out of this project, we intend to produce not only knowledge of robust multivariate time-series clustering but also provide user-friendly packages for the application of our methods."