Discriminant Analysis

Rats were trained to discriminate amphetamine, 1 mg/kg given intraperitoneally, from saline injection in a two-lever operant drug discrimination task. Pseudoephedrine (a sympathomimetic drug with central and peripheral actions) at doses of 10 mg/kg failed to substitute for amphetamine, at 20 mg/kg partial substitution occurred, while at a 40 mg/kg full substitution was seen. The specificity of the amphetamine cue at the training dose used (1 mg/kg) was shown by the finding that a peripherally acting sympathomimetic drug phenylephrine at doses from 0.2 to 0.8 mg/kg failed to substitute for amphetamine. The potential for abuse of pseudoephedrine administered at high doses is discussed.

Ascites can be defined as the abnormal accumulation of fluid inside the peritoneal cavity. Earlier ascites was classified as transudative and exudative based on the total protein concentration of the ascitic fluid. The traditional concept of high pro-tein ascites(> 2.5 g/dl) as exudate ...

The primary purpose of this study was to investigate the strength and anthropometric characteristics of elite and semi-elite tenpin bowlers as compared to non-bowlers, and to establish whether it was possible to discriminate playing level through selected predictor variables. Elite and semi-elite bowlers were distinguished by their bowling score average (BS ave), with participants scoring 200 pin falls and above assigned to the elite group. Eighteen elite bowlers (M=10, F=8; BS ave 213.2± 6.80), 12 semi-elite bowlers (M=7, F=5; BS ave 181.3± 9.36) and 33 sedentary university students (M=14, F=19) were recruited. Anthropometric measurements were taken and isometric arm strength was recorded. Between-group differences were identified through a two-way ANOVA, while discriminant analysis was used to predict group membership. For anthropometric characteristics, the results indicated that the elite bowlers were heavier, had longer lower leg and hand length and had a wider arm span as compared to the non-bowlers. The elite group also had stronger forearm/wrist internal rotation compared to non-bowlers. In addition, the male elite group were stronger than the non-bowling group for arm flexion. There appeared to be a 54% success rate for predicting group membership from selected anthropometric and strength discriminating variables, with forearm/wrist internal rotation strength being the best discriminating variable. It is suggested that coaches should benefit by selecting larger built bowlers with long limbs. Furthermore, bowlers could potentially gain by paying extra attention to increasing the strength levels of forearm/wrist internal rotators and arm flexors during training. These findings also provide normative data for elite and semi-elite bowlers.

http://www.scopus.com/inward/record.url?eid=2-s2.0-84859256312&partnerID=40&md5=e43d03422723c357d50a985754a14880

Phenolic compounds are major constituents of red wines and impact on certain wine-quality parameters. This study has aimed to increase the knowledge base on the phenolic composition of South African red wines by HPLC quan-tification of 39 individual phenols and 2 polymeric, phenolic groups in 260 South African wines of the cultivars Pinotage, Shiraz and Cabernet Sauvignon. Statistical analysis of the collected data revealed significant cultivar dif-ferences in the levels of many of the analysed compounds, as well as in the ratios in which specific anthocyanins are found in the wines. These data may be helpful in the cultivar authentication of wines of these cultivars. Discriminant analysis of the data showed statistically significant separations of cultivars, based on their polyphenol composition, and also how the use of data from a specific wine-producing area or vintage improved the possibilities for success-ful authentication. The data were collected from both pure cultivar wines...

We propose and discuss improved Bayes rules to discriminate between two populations using ordered predictors. To address the problem we propose an alternative formulation using a latent space that allows to introduce the information about the order in the theoretical rules. The rules are first defined when the marginal densities are fully known and then under normality when the parameters are unknown and training samples are available. Several numerical examples and simulations in the paper illustrate the methodology and show that the new rules handle the information appropriately. We compare the new rules with the classical Bayes and Fisher rules in these examples and we show that the misclassification probability is smaller for the new rules. The method is also applied to data from a diabetes study where we again show that the new rules improve over the usual Fisher rule.

A novel approach that combines information provided by the metabolism of pteridines and that of phenylalanine has been applied to the detection of heterozygotes for phenylketonuria. Phenylalanine, tyrosine, biopterin and neopterin have been measured in serum from normal controls and heterozygotes for classical phenylketonuria, before and after a phenylalanine oral load. Significant differences in neopterin and biopterin mean values in fasting serum and in the mean increase of biopterin induced by the phenylalanine load were found between groups. Inclusion of pteridine data in the discriminant analysis significantly improved the resolution of the classical phenylalanine loading test for the detection of heterozygotes for phenylketonuria.