Short article submitted to ECAL 2017 (hosted in Lyon) during my internship at ISIR Lab.
It relates to unsupervized effect clusterization and proposes a clustering method to automatically build a limited set of effects for robots to learn affordances.