Learning from data with structured missingness

Missing data are an unavoidable complication in many machine learning tasks. When data are ‘missing at random’ there exist a range of tools and techniques to deal with the issue. However, as machine learning studies become more ambitious, and seek to learn from ever-larger volumes of heterogeneous data, an increasingly encountered problem arises in which missing values exhibit an association or structure, either explicitly or implicitly. Such ‘structured missingness’ raises a range of challenges that have not yet been systematically addressed, and presents a fundamental hindrance to machine learning at scale. Here we outline the current literature and propose a set of grand challenges in learning from data with structured missingness. Gathering big datasets has become an essential component of machine learning in many scientific areas, but it is unavoidable that some data values are missing. An important and growing effect that needs careful attention, especially when heterogeneous data sources are combined, is that of structured missingness, where data values are missing not at random, but with a specific structure.