Assessing predictability of environmental time series with statistical and machine learning models.
Matthew Bonas, Abhirup Datta, Christopher K Wikle, Edward L Boone, Faten S Alamri, Bhava Vyasa Hari, Indulekha Kavila, Susan J Simmons, Shannon M Jarvis, Wesley S Burr, Daniel E Pagendam, Won Chang, Stefano Castruccio
Author Information
Matthew Bonas: Dept. of Applied and Computational Mathematics and Statistics, University of Notre Dame, Notre Dame, Indiana, USA. ORCID
Abhirup Datta: Dept. of Biostatistics, Johns Hopkins University, Baltimore, Maryland, USA.
Christopher K Wikle: Dept. of Statistics, University of Missouri, Columbia, Missouri, USA. ORCID
Edward L Boone: Dept. of Statistical Sciences and Operations Research, Virginia Commonwealth University, Richmond, Virginia, USA. ORCID
Faten S Alamri: Dept. of Mathematical Sciences, College of Science, Princess Nourah Bint Abdulrahman University, Riyadh, Saudi Arabia.
Bhava Vyasa Hari: Wipro Limited, Bengaluru, India.
Indulekha Kavila: School of Pure and Applied Physics, Mahatma Gandhi University, Kottayam, India. ORCID
Susan J Simmons: Institute for Advanced Analytics, North Carolina State University, Raleigh, North Carolina, USA.
Shannon M Jarvis: Dept. of Mathematics, Trent University, Peterborough, Ontario, Canada. ORCID
Wesley S Burr: Dept. of Mathematics, Trent University, Peterborough, Ontario, Canada. ORCID
Daniel E Pagendam: CSIRO Data61, Eveleigh, Brisbane, Australia.
Won Chang: Div. of Statistics and Data Science, University of Cincinnati, Cincinnati, Ohio, USA. ORCID
Stefano Castruccio: Dept. of Applied and Computational Mathematics and Statistics, University of Notre Dame, Notre Dame, Indiana, USA. ORCID
The ever increasing popularity of machine learning methods in virtually all areas of science, engineering and beyond is poised to put established statistical modeling approaches into question. Environmental statistics is no exception, as popular constructs such as neural networks and decision trees are now routinely used to provide forecasts of physical processes ranging from air pollution to meteorology. This presents both challenges and opportunities to the statistical community, which could contribute to the machine learning literature with a model-based approach with formal uncertainty quantification. Should, however, classical statistical methodologies be discarded altogether in environmental statistics, and should our contribution be focused on formalizing machine learning constructs? This work aims at providing some answers to this thought-provoking question with two time series case studies where selected models from both the statistical and machine learning literature are compared in terms of forecasting skills, uncertainty quantification and computational time. Relative merits of both class of approaches are discussed, and broad open questions are formulated as a baseline for a discussion on the topic.
