On the consistency of the blocked neural network estimator in time series analysis.

We describe a nonlinear regression problem, where the regression functions have an additive structure and the dependent variable is a one-dimensional time series. Multivariate time series with unknown time delay operators are used as independent variables. By fitting a feedforward neural network with block structure to the data, we estimated the additive regression function and, parallel to this, the time lags. We present the consistency proof of neural network weights estimator and the time lag estimator independently from each other. In the practical part of the article, we present the useful feature of blocked neural networks to estimate the relevance measures of each input variable in a simple way. Furthermore, we propose an approach to solve the well-known variable selection problem for the class of nonlinear multivariate beta-mixing time series models considered here. Finally, we apply the methodology to an artificial example.

Quality of organically and conventionally grown potatoes: four-year study of micronutrients, metals, secondary metabolites, enzymic browning and organoleptic properties.

The quality of potatoes from organic and conventional farming was investigated in this study. Tubers of eight potato varieties, organically and conventionally produced at one or two geographical sites in controlled field trials, were collected in four consecutive harvests from 1996-1999. The parameters analysed included nitrate, trace elements (As, Cd, Co, Cu, Fe, Hg, Mn, Ni, Pb, Se, Zn), vitamin C, potato glycoalkaloids, as well as chlorogenic acid, polyphenol oxidase and rate of tuber enzymic browning. The results indicated lower nitrate content and higher vitamin C and chlorogenic acid content to be the parameters most consistently differentiating organically from conventionally produced potatoes. Elevated concentrations of glycoalkaloids were also observed throughout the experiments in some potato varieties grown in organic farming systems. Principal component analysis (PCA) of the analytical and other data using three PCs confirmed a good separation between the organically and conventionally produced potatoes when studied in single crop years. However, score-plots (objects) and loading-plots (variables) of pooled results from the consecutive harvests showed that between the years' changes and also variety as well as geographical variations are equally or more important factors determining the quality of potatoes than the farming system. Further studies of various marker compounds of potato quality related to the organic or conventional farming systems should be performed before unbiased information can be given to the consumers.