Truth in complex adaptive systems models should be based on truth by constructive verification

It is argued that the truth status of emergent properties of complex adaptive systems models should be based on an epistemology of proof by constructive verification and therefore on the ontological axioms of a non-realist logical system such as constructivism or intuitionism. ‘Emergent’ properties of complex adaptive systems (CAS) models create particular epistemological and ontological challenges. These challenges bear directly on current debates in the philosophy of mathematics and in theoretical computer science. CAS research, with its emphasis on computer simulation, is heavily reliant on models which explore the entailments of Formal Axiomatic Systems (FAS). The incompleteness results of Gödel, the incomputability results of Turing, and the Algorithmic Information Theory results of Chaitin, undermine a realist (platonic) truth model of emergent properties. These same findings support the hegemony of epistemology over ontology and point to alternative truth models such as intuitionism, constructivism and quasi-empiricism.

Exopolysaccharides produced by Bifidobacterium longum IPLA E44 and Bifidobacterium animalis subsp lactis IPLA R1 modify the composition and metabolic activity of human faecal microbiota in pH-controlled batch cultures

Exopolysaccharides (EPS) isolated from two Bifidobacterium strains, one of human intestinal origin (Bifidobacterium longum subsp. longum IPLA E44) and the other from dairy origin (Bifidobacterium animalis subsp. lactis IPLA R1), were subjected to in vitro chemically simulated gastrointestinal digestion. which showed the absence of degradation of both polymers in these conditions. Polymers were then used as carbon sources in pH-controlled faecal batch cultures and compared with the non-prebiotic carbohydrate glucose and the prebiotic inulin to determine changes in the composition of faecal bacteria. A set of eight fluorescent in situ hybridisation oligonucleotide probes targeting 16S rRNA sequences was used to quantify specific groups of microorganisms. Growth of the opportunistic pathogen Clostridium histolyticum occurred with all carbohydrates tested similarly to that found in negative control cultures without added carbohydrate and was mainly attributed to the culture conditions used rather than enhancement of growth by these substrates. Polymers E44 and RI stimulated growth of Lactobacillus/Enterococcus, Bifidobacterium, and Bacteroides/Prevotella in a similar way to that seen with inulin. The EPS RI also promoted growth of the Atopobium cluster during the first 24 h of fermentation. An increase in acetic and lactic acids was found during early stages of fermentation (first 10-24 h) correlating with increases of Lactobacillus, Bifidobacterium, and Atopobium. Propionic acid concentrations increased in old cultures, which was coincident with the enrichment of Clostridium cluster IX in cultures with EPS RI and with the increases in Bacteroides in cultures with both microbial EPS (RI and E44) and inulin. The lowest acetic to propionic acid ratio was obtained for EPS E44. None of the carbohydrates tested supported the growth of microorganisms from Clostridium clusters XIVa+b and IV, results that correlate with the poor butyrate production in the presence of EPS. Thus, EPS synthesized by bifidobacteria from dairy and intestinal origins can modulate the intestinal microbiota in vitro, promoting changes in some numerically and metabolically relevant microbial populations and shifts in the production of short chain fatty acids. (C) 2009 Elsevier B.V. All rights reserved.

In vitro fermentation of carbohydrates by porcine faecal inocula and their influence on Salmonella Typhimurium growth in batch culture systems

The aim of this study was to evaluate in vitro the influence of fermentable carbohydrates on the activity of porcine microbiota and survival of Salmonella Typhimurium in a batch culture system simulating the porcine hindgut. The carbohydrates tested were xylooligosaccharides, a mixture of fructooligosaccharides/inulin (FIN), fructooligosaccharides (FOS), gentiooligosaccharides (GEO) and lactulose (LAC). These ingredients stimulated the growth of selected Bifidobacterium and Lactobacillus species in pure cultures. In batch cultures, the carbohydrates influenced some fermentation parameters. For example, GEO and FIN significantly increased lactic acids compared with the control (no added carbohydrate). With the exception of LAC, the test carbohydrates increased the production of short-chain fatty acid (SCFA) and modified SCFA profiles. Quantitative analysis of bacterial populations by FISH revealed increased counts of the Bifidobacterium group compared with control and, with exception of FOS, increased Lactobacillus, Leuconostoc and Weissella spp. counts. Salmonella numbers were the lowest during the fermentation of LAC. This work has looked at carbohydrate metabolism by porcine microbiota in a pH-controlled batch fermentation system. It provides an initial model to analyse interactions with pathogens.

Synthesis of galacto-oligosaccharide from lactose using beta-galactosidase from Kluyveromyces lactis: studies on batch and continuous UF membrane-fitted bioreactors

A study of galacto-oligosaccharides (GOS) synthesis from lactose with beta-galactosidase from Kluyveromyces lactis (Maxilact(R) L2000) was carried out. The synthesis was performed using various initial lactose concentrations ranging from 220 to 400 mg/mL and enzyme concentrations ranging from 3 to 9 U/mL, and was investigated at 40degreesC and pH 7, in a stirred-tank reactor. In the experimental range examined, the results showed the amount of GOS formed depended on lactose concentration but not on enzyme concentration. Galactose was a competitive inhibitor, while glucose was a non-competitive inhibitor. In a further study, a laboratory-scale reactor system, fitted with a 10-kDa NMWCO composite regenerated cellulose membrane, was used in a continuous process. The reactor was operated in cross-flow mode. The effect of operating pressures on flux and productivity was investigated by applying different transmembrane pressures to the system. The continuous process showed better production performance compared to the batch synthesis with the same lactose and enzyme concentrations at 40degreesC, pH 7. Comparison of product structures from batch and continuous processes, analyzed by HPAEPAD and methylation analysis, showed similarities but differed from the structures found in a commercial GOS product (Vivinal(R)GOS). (C) 2004 Wiley Periodicals, Inc.

Verification and FPGA circuits of a block-2 fast path-based predictor

This paper formally derives a new path-based neural branch prediction algorithm (FPP) into blocks of size two for a lower hardware solution while maintaining similar input-output characteristic to the algorithm. The blocked solution, here referred to as B2P algorithm, is obtained using graph theory and retiming methods. Verification approaches were exercised to show that prediction performances obtained from the FPP and B2P algorithms differ within one mis-prediction per thousand instructions using a known framework for branch prediction evaluation. For a chosen FPGA device, circuits generated from the B2P algorithm showed average area savings of over 25% against circuits for the FPP algorithm with similar time performances thus making the proposed blocked predictor superior from a practical viewpoint.

Face verification using Gabor wavelets and AdaBoost

This paper presents a new face verification algorithm based on Gabor wavelets and AdaBoost. In the algorithm, faces are represented by Gabor wavelet features generated by Gabor wavelet transform. Gabor wavelets with 5 scales and 8 orientations are chosen to form a family of Gabor wavelets. By convolving face images with these 40 Gabor wavelets, the original images are transformed into magnitude response images of Gabor wavelet features. The AdaBoost algorithm selects a small set of significant features from the pool of the Gabor wavelet features. Each feature is the basis for a weak classifier which is trained with face images taken from the XM2VTS database. The feature with the lowest classification error is selected in each iteration of the AdaBoost operation. We also address issues regarding computational costs in feature selection with AdaBoost. A support vector machine (SVM) is trained with examples of 20 features, and the results have shown a low false positive rate and a low classification error rate in face verification.