My research in statistics has both a methodological and an applied component. In the past I have worked on data analysis in cosmology,
statistical signal processing in optics, and robust spectrum estimation in geophyiscs.
My current research is focused upon neurostatistics - the developement and application of statistical methodology to problems in the neurosciences.
- Inferring Evoked Connectivity by Adaptive Perturbation - The aim of this work is to provide statistically principled
methodology for the purposes of understanding cortical connectivity in a real-time
experimental paradigm involving neural stimulation.
- Statistical Inference with Point Process Models of Neural Spike Trains - For example, investigating and applying principled approaches of separating confounds within a point process/generalized linear modeling paradigm of analyzing neural spike train data.
- Spike-Field Association - Developing statistical methodology appropriate for
modeling local-field potential interactions with single-neuron spiking activity. This work combines parametric point process modeling
with classical non-parametric spectral analysis.
- Time Series Analysis - Topics include: quadratic inverse spectrum estimation, phase estimation, frequency-dependent measures of association such as the phase-locking value (PLV), and robust electroencephalography (EEG) referencing.