Time dependent neural network models for detecting changes of state in complex processes: applications in earth sciences and astronomy.

A computational intelligence approach is used to explore the problem of detecting internal state changes in time dependent processes; described by heterogeneous, multivariate time series with imprecise data and missing values. Such processes are approximated by collections of time dependent non-linear autoregressive models represented by a special kind of neuro-fuzzy neural network. Grid and high throughput computing model mining procedures based on neuro-fuzzy networks and genetic algorithms, generate: (i) collections of models composed of sets of time lag terms from the time series, and (ii) prediction functions represented by neuro-fuzzy networks.

Comparison of instrumental neutron activation analysis of geological materials with other multielement techniques.

Results obtained for 12 elements in approximately 1600 rocks by instrumental neutron activation analysis (INAA) are compared with those obtained by ICP emission spectrometry (ICP-ES), XRF, and atomic-absorption spectrometry (AAS). Sample duplicates and two controls are used to evaluate the precison of the methods investigated. Application of a method (Maximum Likelihood Functional Relationship) to determine and quantify rotational and translational bias is demonstrated.