Explaining Effective Learning by Analogical Reasoning

Machine learning algorithms are usually considered as explicit learning strategies requiring large data samples. Contrary to these accounts, cognitive learning seems to be based on significantly less amounts of training data and occurs often in the form of implicit learning. In order to close this gap we propose to explain these discrepancies by a form of analogical learning, bridging these two traditions. Using heuristic-driven theory projection (HDTP) as the framework for analogy making we can productively model learning aspects with sparse training data.