About my work
Currently I am a Research Fellow at The University of Nottingham working within the RootStruc project led by Professors Murray Lark and Karl Ritz which aims to understand how and until what extent roots can influence soil structure. To address these questions I am proposing statistical methods based on smoothing splines and wavelets to model volumes (3D representations) and images (2D) from X-ray Computarized Tomography (CT) of rooted soil. For each experimental unit, the CT scan produces a volume corresponding to millions of data to analyse. Therefore this is a challenging big data problem which requires parallel computing and novel statistical methods.
I have a PhD in Statistics from The University of Manchester where I worked on “Smoothing splines with unknown change-points in responses for longitudinal and spatial data” supervised by Professor Jianxin Pan. Among others, the key novel ideas of this work are:
- Multidimensional smoothing and its approximation using additive models.
- Interpolation methods (smoothing) for dependent data (longitudinal, time series, spatial).
- Estimation of curves and surfaces when unknown change-points or contours of change are present.
- Estimation of these models based on bayesian and maximum likelihood approaches using mixed models.
- Tests for linearity and planarity.