Nonparametric inference, dependence, high dimension, directional data, incomplete data, set estimation. The advances in nonparametric inference on curves have provided answers to applied problems in medicine, biology, economics or environmental sciences, among other fields. In all these contexts, data recording mechanisms have evolved in such a way that the current scenario shows huge amounts of data (“Big data”). These data usually come from stochastic systems (DGS: data generating system) which may be described as complex, understading as “complex systems” those DGSs whose response or design is far from the classical inferential setting (real and independent observations, as the simplest case). We may distinguish DGSs which are complex because of the data nature: functional or high dimensional data, directional data, imcomplete data (missing, censored, truncated), images or sets. Or DGSs which are complex due to underlying structures and dynamics: spatial and/or temporal dependent data; DGSs which are modelled through “complex” regressión with random effects, structured or quantile models. Or in general, any combination of non-standard data (even from combined nature, such as directional and high dimensional, for instance) and complex dynamics.

Innpar2D is born with the aim of obtaining methodological advances which allow to describe and characterize the behaviour of complex DGSs. The project is structured in two blocks: (B1) nonparametric inference and (B2) exploratory methods, software and applications. Working methodology is focused on theoretical developments, validation through simulation studies, software and exploratory methods production (and their disposal for the scientific community) and the consideration of practical problems in fields such as medicine, environmental sciences, actuarial sciences and lingüistics.

The research group for Innpar2D has been designed with the goal of keeping the excelence levels reached in recent years, with an active dissemination of research results; training young researchers and strengthening the relation with bussinesses and industry, with new collaborations for technology transfer. Our scientific foregoing guarantees the success of this new proposal.

This proyect is a continuation from MTM2013-41383-P (Innpar), more details in http://eio.usc.es/pub/innpar/.