Sparse-filtering is an unsupervised feature learning algorithms which explicitly optimizes for sparse activation of the generated feature extractors. This notebook illustrates the algorithm on the Olivetti dataset.
Reliability diagrams are useful for checking if the predicted probabilities of a binary classifier are well calibrated. For perfectly calibrated predictions, the curve in a reliability diagram should be as close as possible to the diagonal/identity. This post compares the reliability diagrams of different classfiers on artifical data and discusses the respective pros and cons.