ABSTRACT
This chapter motivates OODA in the context of the rapidly changing fields of statistics, data science, and data analytics. A wide range of Big Data contexts, including each of the dimension and sample size being large are considered. While Big Data is acknowledged as a major challenge, an important point is that an even greater challenge is Complex Data. Key concepts such as data visualization based on modes of variation are illustrated using the curves as data objects, i.e. Functional Data Analysis, example of the Spanish Mortality data. The analysis links with various societal trends in a number of interesting ways. Another important aspect of OODA is non-Euclidean data objects, as typified by the Bladder-Prostate-Rectum image data set. Various shape representations, which result in data objects naturally lying on curved manifolds, including the skeletal approach are discussed. A Bayesian approach to particularly challenging segmentation problems is seen to be very effective.
