Change of representation and inductive bias (original) (raw)

The document discusses the significance of representation and inductive bias in machine learning. It emphasizes that the choice of language and representation affects the effectiveness of learning algorithms, highlighting the trade-offs between broad and narrow hypothesis spaces. It recounts the First International Workshop on Change of Representation and Inductive Bias held in 1988, which brought together diverse researchers to explore related topics such as constructive induction and problem reformulation. The proceedings include various papers that evolved from the discussions at this workshop.