• Title of article

    Least third-order cumulant method with adaptive regularization parameter selection for neural networks

  • Author/Authors

    Leung، Chi-Tat نويسنده , , Chow، Tommy W.S. نويسنده ,

  • Issue Information
    روزنامه با شماره پیاپی سال 2001
  • Pages
    -168
  • From page
    169
  • To page
    0
  • Abstract
    The novel extractants, six diglycolamides synthesized in our laboratory, were investigated for actinide extraction from nitric acid into n-dodecane. The dependence of the distribution coefficients of Eu(lll) and Am(lll) on the length of alkyl chain in the extractants was examined. Among the diglycolamides studied, N,N,Nʹ,Nʹ-tetraoctyl-3-oxapentanediamide (TOOPDA) and N,N,Nʹ,Nʹ -tetradecyl-3-oxapentanediamide (TDOPDA) showed a sufficient solubility in n -dodecane, due to their appropriate lipophilicity modified by increasing the length of the alkyl chain attached to amidic N atoms. The distribution coefficients, DM, of actinides obtained by TOOPDA increased with an increase in HNO3 concentration. The number of diglycolamide molecules coordinated to the actinide ions was estimated to be three for Th(IV), U(VI), Pu(IV), and four for Am(III) and Cm(III) by slope analysis. The order of DM for the actinides at high nitric acid concentrations is An(III), An(IV) > An(VI) > An(V) and that for lanthanides, heavier > lighter. It is concluded that TOOPDA diluted in n-dodecane can extract both actinides and lanthanides completely from aqueous HNO3 solution.
  • Keywords
    Third-order cumulant , NEURAL NETWORKS , regularization , Generalization capability
  • Journal title
    ARTIFICIAL INTELLIGENCE (NON MEMBERS) (AI)
  • Serial Year
    2001
  • Journal title
    ARTIFICIAL INTELLIGENCE (NON MEMBERS) (AI)
  • Record number

    48107