Isotonic regression is a method for obtaining a monotonic fit for 1-dimensional data. Let’s say we have data such that . (We assume no ties among the ‘s for simplicity.) Informally, isotonic regression looks for such that the ‘s approximate … Continue reading →