Science: similarities and differences among catalysts: Clustering and profiling diverse data on chemical reactions

  • Authors:

  • Raúl E. Valdés-Pérez;Andrew V. Zeigarnik;Jérôme Pesenti

  • Affiliations:

  • Senior Research Computer Scientist, Computer Science Department, Carnegie Mellon University, Pittsburgh, Pennsylvania;Research Scientist, Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences, Moscow;Visiting Professor, Computer Science Department, Carnegie Mellon University, Pittsburgh, Pennsylvania

  • Venue:
  • Handbook of data mining and knowledge discovery
  • Year:
  • 2002

Quantified Score

Hi-index 0.00

Visualization

Abstract

The science of chemistry has seen less recent use of knowledge discovery techniques than sister sciences like biology, possibly because the comparatively modest size of typical chemical data sets do not obviously call out for data mining. The mistaken impression that small-to-medium sized data sets need at best simple methods is widespread. In this article we apply a Combination of classical hierarchical clustering (Hartigan, 1975) and profiling methods (Valdes-Perez et al., 2000) to gain insight into the respective capabilities of different metals to catalyze important reactions in chemistry (see Chapter 16.5.2 in this handbook for further background on clustering methods). As far as we know, this is one of the very few applications of data mining within the broad subfield of chemistry known as catalysis.