The objective of this project is to revive the study of stylistic development of painting by developing a visual analytics framework combining subjective connoisseurship with computer learning. Art history experts have the unique ability to make detailed assessments of small sets of images, but the vast amounts of images available make it impossible for individuals to thoroughly analyse them all. This project will train computers to become objective connoiseurs for attribution and authentication of single art works with the help of subjective human connoiseurs. We will research how image features can be optimized by incorporating knowledge on the creation process and characteristics of materials. From there, we will consider how to visualize the results in such a way that art historians can interactively annotate and re-group the paintings, so in a sense ‘correct’ the machine output where necessary. Machine learning methods will then be developed which use those interactions to improve the model the system uses to define similarities in paintings. Integrating the resulting data-driven similarities with structured metadata that already exist for the paintings will yield stylistic histories in terms of visual characteristics, place, and time. The resulting tools will be made available as open source software.