Research

Home     Research     Publications      Teaching      Contact

 

My research is aimed at developing statistical models for time series forecasting.

Currently, I am working with Dr Max Little on ubiquitous, inexpensive non-invasive technologies for objective detection and monitoring of Parkinson's Disease (PD) symptoms. My work involves analysing data for: 1) Speech, 2) Gait, 3) Posture, 4) Finger tapping, and, 5) Reaction times, recorded using smartphones. The aim of this work is to find discriminatory patterns in the data which can help distinguish PD participants from healthy controls.

 

Application areas :-

  •  Healthcare: a) Detecting, Monitoring and Predicting the Symptoms of Parkinson’s Disease/Friedreich's Ataxia using High-frequency Datasets Collected via Smartphones, and, b) Predicting Arrivals, Admissions and Discharges Across Hospitals in the West Midlands

  •  Energy: Modelling, a) Total National Electricity Demand for Great Britain and France with Emphasis on Anomalous Periods, and, b) Electricity Consumption for Residential Consumers and SMEs using the Irish Smart Meter Data    

  •  Macroeconomics: Generating Probability Density Forecasts for the US Gross National Product (GNP) using Novel Nonlinear and Nonparametric Models

  •  Climate: Comparing Forecasts from Global Climate Models with Statistical Non- linear Models Using Time Series for the Global Surface Air Temperature