Multivariate time-series segmentation for immunological data analysis

Developing new methodologies for (biomedical, clinical, immunological, molecular) data analysis is a main research interest in both bioin- formatics and computational biology. In particular, efficient and fast computational methods are proposed in the literature of machine learning (for data clustering, and feature extraction) to infer new knowledge from given data. A vivid interest is particularly devoted to immunological research, as most of the human deseases are induced by some fall or misplay in our body defence system. More specifically, a recent broad interest is focused on the lifetime aging of immune system, in terms of changes of immune mechanisms of an individual during his/her infancy, growing age, mature age, and senescence. Indeed, an age-related decline, referred to as immunosenescence, seems characterized by a decrease in cell-mediated immune functions, where defects in T- and B-cell functions coexist, and has social/commercial impact related to vaccination in elderly persons. Here we present a couple of different models respectively based on MP systems and on piecewise linear segmentation. Moreover, we discuss our current work on an immunological dataset analysis, we are approaching by a combination of time-series clustering and segmentation methodologies.