Statistical methods, such as independent component analysis, have been successful in learning local low-level features from natural image data. Here we extend these methods for le...
for data abstraction and justifies reasoning by simulation. Representation independence has been shown for a variety of languages and constructs but not for shared references to mu...
Abstract. This paper compares two grammar based Evolutionary Automatic Programming methods, Grammatical Evolution (GE) and Chorus. Both systems evolve sequences of derivation rules...
We take another look at the general problem of selecting a preferred probability measure among those that comply with some given constraints. The dominant role that entropy maximi...