作者: Laurent A. Baumes , Pierre Collet
DOI: 10.1016/J.COMMATSCI.2008.03.051
关键词:
摘要: The strong feature dependencies that exist in catalyst description do not permit using common algorithms while loosing crucial information. Data treatments are restricted by the form of input data making full use experimental information impossible, confining experimentation studies, and reducing one primary goals HTE: to enlarge search space. Consequently, an advanced representation catalytic supporting intrinsic complexity heterogeneous structure is proposed. Likewise, optimization strategy can manipulate efficiently such type, permitting a valuable connection between algorithms, high-throughput (HT) apparatus, databases, depicted. Such new methodology enables integration domain knowledge through its configuration considering study be investigated. For first time catalysis, conceptual examination genetic programming (GP) achieved.