Mean Field Theory Neural Networks for Feature Recognition,Content Addressable Memory and Optimization

Various applications of the mean field theory (MFT) technique for obtaining solutions close to optimal minima in feedback networks are reviewed. Using this method in the context of the Boltzmann machine gives rise to a fast deterministic learning algorithm with a performance comparable with that of the backpropagation algorithm (BP) in feature recognition applications. Since MFT learning is bidirectional its use can be extended from purely functional mappings to a content addressable memory. The storage capacity of such a network grows like O (10–20)n_H with the number of hidden units. The MFT learning algorithm is local and thus it has an advantage over BP with respect to VLSI implementations. It is also demonstrated how MFT and BP are related in situations where the number of input units is much larger than the number of output units. In the context of-finding good solutions to difficult optimization problems the MFT technique again turns out to be extremely powerful. The quality of the solutions for large travelling salesman and graph partition problems are in parity with those obtained by optimally tuned simulated annealing methods. The algorithm employed here is based on multistate K-valued (K > 2) neurons rather than binary (K = 2) neurons. This method is also advantageous for more nested decision problems like scheduling. The MFT equations are isomorfic to resistance-capacitance equations and hence naturally map onto custom-made hardware. With the diversity of successful application areas the MFT approach thus constitutes a convenient platform for hardware development.     

Sorbent Extraction and Analysis of Volatile Metabolites Synthesized by Lactic Acid Bacteria in a Synthetic Medium

Volatile metabolites synthesized by strains of Luctobacillus brevis, L. fructivorans, L. hilgardii, L. plantarum, Pediococcus parvulus, and Leuconostoc oenos, originally isolated from commercial wines, were extracted from a synthetic medium using C₁₈ sorbent cartridges and analyzed using gas-liquid chromatography. All organisms synthesized isoamyl alcohol, 1-hexanol, 3-(methylthio)-1-propanol, and 2-phenyl ethanol in varying concentrations. Several other alcohols, acids, aldehydes, and esters were produced or utilized by some species or strains and enabled differentiation of the organisms. Recoveries of metabolites were dependent on the compounds, matrix, and presence of ethanol.     

EFFECTS OF CULTURAL CONDITIONS ON THE SYNTHESIS OF SALMONELLA TOXIN

The influence of culture age, pH, aeration, and temperature on Salmonella toxin production was investigated using fermentor cultures containing casamino acids-yeast extract as a growth medium. About 75% or more of total Salmonella toxin was synthesized and released when the cultures were harvested during stationary phase of growth. Increased production and release of the toxin was also observed when the cultures were regulated at pH 6.0, 7. 0, or under alkaline control. The optimal temperature for the synthesis of Salmonella toxin was at 37°C; increased aeration of the cultures (500 rpm) appeared to enhance Samonella toxin production.     

Na+ Binding as Measured by 23Na Nuclear Magnetic Resonance Spectroscopy Influences the Perception of Saltiness in Gum Solutions

Binding of Na⁺ in aqueous gum systems as determined by ²³Na nuclear magnetic resonance (NMR) spectroscopy and its relations to perceived saltiness were examined. Two levels of NaCl (0.1% and 0.2%) were added to two concentrations (0.1% and 0.3%) of two ionic (xanthan and kappa carrageenan) and two non-ionic (locust bean and guar) gum solutions. Saltiness perception was affected by the ionic properties of the gums. NMR transverse relaxation rates (R₂, see^?l) indicated Na⁺ was less mobile in ionic than nonionic systems. Ionic gums correspondingly suppressed saltiness perception- compared to nonionic gums. As Na⁺ increased in both ionic and nonionic systems, R₂ values converged and perceived saltiness equalized. Food components that bind Na⁺ may suppress saltiness perception, which may be important in low-sodium foods.     

Reflectance factor dynamics of boreal forest clear-cut communities during early secondary succession

Abstract

Spectral reflectance factors in the visible and near-infrared spectral bands were measured over five years on permanent sample plots in forest clear-cut communities in Estonia. The spectral-temporal profiles of vegetation indices Greenness and Brightness were calculated from discrete measurements of reflectance factors. The following hypotheses concerning successional reflectance dynamics in boreal forest communities following clear-cutting and understory destruction were tested. 1. Changes in the Greenness seasonal profile peak value (PV) during secondary succession are mainly directional and related to time after clear-cutting rather than to environmental fluctuations. 2, The rate of succession, i.e., difference between seasonal PVs of consecutive years, declines over time. 3. The seasonal PVs of forest communities from different site types become more similar (converge) during secondary succession.

The results show that the secondary succession of forests in reflectance terms is directed towards the enhancement of contrast between visible to near-infrared reflectance, towards obtaining maximum spectral vegetation index Greenness value at a given Brightness level. The PV changes during secondary succession were explained by the age of the communities. The seasonal PVs of different site type clear-cuts converge during early secondary succession, resulting from different speed of clear-cut recolonization on fertile and poor sites.