How to estimate and forecast power consumption patterns using data time series from geographically divers measurement stations. That’s the main question that. Gavin Shaddick, reader in Statistics at University of Bath Department of Mathematical Sciences, England, addressed in the course Detecting patterns in space and time: statistical analysis of big data in the power industry, held at Universidad de Chile Center for Mathematical Modeling (CMM) on January 12 and 13, 2016.
Shaddick, who also regularly works on medical and environmental data analysis with the World Health Organization (WHO) described data reduction techniques that allow information required by industries to be retained in smaller, more manageable datasets. The idea is using the information to infer and predict consumer behavior and optimizing the distribution of energy in different geographical areas.
The course also included practical sessions, given by Bath University research engineer Amelia Jobling. She explained how to use these statistical methods in different applied areas, which was highly appreciated by researchers and engineers CMM.
“The course addressed a problem that is very common in our laboratory, such as treating sets of time series with the difficulty of also involving spatial dimensions. The course showed an outline of methods to find solutions using R programming language. It was very precise,” said CMM project engineer Gonzalo Ríos.
For the project engineer Jorge Prado, this course “serve to match our knowledge and step up our research. Here we could also talk about what we do in our laboratories and share our experiences.”
The course is part of the new scientific partnership between CMM and University of Bath Department of Mathematical Sciences.
