Collaborative Filtering for Brain-Computer Interaction Using Transfer Learning and Active Class Selection

Brain-computer interaction (BCI) and physiological computing are terms that refer to using processed neural or physiological signals to influence human interaction with computers, environment, and each other. A major challenge in developing these systems arises from the large individual differences typically seen in the neural/physiological responses. As a result, many researchers use individually-trained recognition algorithms to process this data. In order to minimize time, cost, and barriers to use, there is a need to minimize the amount of individual training data required, or equivalently, to increase the recognition accuracy without increasing the number of user-specific training samples. One promising method for achieving this is collaborative filtering, which combines training data from the individual subject with additional training data from other, similar subjects. This paper describes a successful application of a collaborative filtering approach intended for a BCI system. This approach is based on transfer learning (TL), active class selection (ACS), and a mean squared difference user-similarity heuristic. The resulting BCI system uses neural and physiological signals for automatic task difficulty recognition. TL improves the learning performance by combining a small number of user-specific training samples with a large number of auxiliary training samples from other similar subjects. ACS optimally selects the classes to generate user-specific training samples. Experimental results on 18 subjects, using both nearest neighbors and support vector machine classifiers, demonstrate that the proposed approach can significantly reduce the number of user-specific training data samples. This collaborative filtering approach will also be generalizable to handling individual differences in many other applications that involve human neural or physiological data, such as affective computing.     

Geographic variation in reproduction between two populations of the bonnethead shark,Sphyrna tiburo

Synopsis A study of two populations of the bonnethead shark,Sphyrna tiburo was conducted in Florida Bay and Tampa Bay, Florida from September 1982 to December 1986. The maintenance of sharks in captivity at the Marine Science and Conservation Center in the Florida Keys, and the collection of sharks from widely separated geographical areas allowed the examination of latitudinal variation in reproduction. Several reproductive parameters were found to differ: (1) size at maturation, (2) age at maturation, (3) time of fertilization, (4) rate of embryonic development, (5) size at birth, (6) the energetic investment in producing offspring, (7) gestation period, and (8) the incidence of infertility. Average litter size and maximum age of females was not different between the two populations. These contrasting life history parameters are not easily explained. Food limitation and seasonal differences between the two areas are considered as factors controlling reproduction in these populations. It is noteworthy that the average size of adult females in Tampa Bay is significantly greater than that of Florida Keys sharks. This size difference may be important in explaining the observed differences in reproduction.     

Physical Activity and Physical Fitness in African-American Girls With and Without Obesity

WARD, DIANNE S, STEWART G TROST, GWEN FELTON, RUTH SAUNDERS, MARY ANN PARSONS, MARSHA DOWDA, RUSSELL R PATE. Physical activity and physical fitness in African-American girls with and without obesity. Lack of physical activity and low levels of physical fitness are thought to be contributing factors to the high prevalence of obesity in African-American girls. To examine this hypothesis, we compared habitual physical activity and physical fitness in 54 African-American girls with obesity and 96 African-American girls without obesity residing in rural South Carolina. Participation in vigorous (6 METs) (VPA) or moderate and vigorous physical activity (4 METs) (MVPA) was assessed on three consecutive days using the Previous Day Physical Activity Recall. Cardiorespiratory fitness was assessed using the PWC 170 cycle ergometer test. Upper body strength was determined at two sites via isometric cable tensiometer tests. Relative to their counterparts without obesity, girls with obesity reported significantly fewer 30-minute blocks of VPA (0. 90 ± 0. 14 vs. 1. 3 ± 0. 14) and MVPA (1. 2 ± 0. 18 vs. 1. 7 ± 0. 16) (p<0. 01). Within the entire sample, VPA and MVPA were inversely associated with body mass index (r=?0. 17 and r=?0. 19) and triceps skinfold thickness (r=?0. 19 and r=?0. 22) (p<0. 05). In the PWC 170 test and isometric strength tests, girls with obesity demonstrated absolute scores that were similar to, or greater than, those of girls without obesity; however, when scores were expressed relative to bodyweight, girls with obesity demonstrated significantly lower values (p<0. 05). The results support the hypothesis that lack of physical activity and low physical fitness are important contributing factors in the development and/or maintenance of obesity in African-American girls.     

Focal adhesion kinase (FAK) regulates insulin-stimulated glycogen synthesis in hepatocytes.

Experimental data support a role for FAK, an important component of the integrin signaling pathway, in insulin action. To test the hypothesis that FAK plays a regulatory role in hepatic insulin action, we overexpressed wild type (WT), a kinase inactive (KR), or a COOH-terminal focal adhesion targeting (FAT) sequence-truncated mutant of FAK in HepG2 hepatoma cells. In control untransfected (NON) and vector (CMV2)- and WT-transfected cells, insulin stimulated an expected 54 +/- 13, 37 +/- 4, and 47 +/- 12 increase in [U-(14)C]glucose incorporation into glycogen, respectively. This was entirely abolished in the presence of either KR (-1 +/- 7%) or FAT mutants (0 +/- 8%, n = 5, p < 0.05 for KR or FAT versus other groups), and this was associated with a significant attenuation of incremental insulin-stimulated glycogen synthase (GS) activity. Insulin-stimulated serine phosphorylation of Akt/protein kinase B was significantly impaired in mutant-transfected cells. Moreover, the ability of insulin to inactivate GS kinase-3beta (GSK-3beta), the regulatory enzyme immediately upstream of GS, by serine phosphorylation (308 +/- 16, 321 +/- 41, and 458 +/- 34 optical densitometric units (odu) in NON, CMV2, and WT, respectively, p < 0.02 for WT versus CMV2) was attenuated in the presence of either FAT (205 +/- 14, p < 0.01) or KR (189 +/- 4, p < 0.005) mutants. FAK co-immunoprecipitated with GSK-3beta, but only in cells overexpressing the KR (374 +/- 254 odu) and FAT (555 +/- 308) mutants was this association stimulated by insulin compared with NON (-209 +/- 92), CMV2 (-47 +/- 70), and WT (-39 +/- 31 odu). This suggests that FAK and GSK-3beta form both a constitutive association and a transient complex upon insulin stimulation, the dissociation of which requires normal function and localization of FAK. We conclude that FAK regulates the activity of Akt/protein kinase B and GSK-3beta and the association of GSK-3beta with FAK to influence insulin-stimulated glycogen synthesis in hepatocytes. Insulin action may be subject to regulation by the integrin signaling pathway, ensuring that these growth and differentiation-promoting pathways act in a coordinated and/or complementary manner.     

Trends in savanna structure and composition along an aridity gradient in the Kalahari

Abstract. The Kalahari sand sheet occupies 2.5 million ha in southern Africa. It is an area with relatively similar deep aeolian soils, and a strong south to north gradient in rainfall, from 200 to 1000 mm mean annual precipitation (MAP) in the region studied. This provides an excellent basis for gradient studies at the subcontinental scale. This paper briefly reviews the literature on the vegetation of the Kalahari and describes the vegetation structure and composition at 11 new sites. There is a clear gradient in woody plant biomass (as indexed by basal area) from south to north. Above the minimum level of 200 mm MAP, the woody basal area increases at a rate of ca. 2.5 m².ha^‐1 per 100 mm MAP. Mean maximum tree height also increases along the gradient, reaching 20 m at ca. 800 mm MAP. The number of species to contribute > 95% of the woody basal area increases from one at 200 mm to 16 at 1000 mm MAP. Members of the Mimosaceae (mainly Acacia) dominate the tree layer up to 400 mm MAP. They are replaced by either the Combretaceae (Combretum or Terminalia) or Colophospermum mopane of the Caesalpinaceae between 400 and 600 mm MAP, and by other representatives of the Caesalpinaceae above 600 mm MAP. The vegetation is largely deciduous up to 1000 mm MAP, except for species that apparently have access to groundwater, which may be locally dominant above about 600 mm MAP.     

Intestinal absorption of peptides peptide uptake by small intestine of Rana pipiens

The absorption of glycyl-l-leucine is studied using the small intestine of Rana pipiens perfused through the vascular bed. It is found that the transfer into the vascular bed of glycine and leucine from the peptide in the lumen occurs in part by a process independent from those for the transfer of the free amino acids. This process is particularly effective for the transfer of glycine.     

Developmental variability and the limits of adaptation: Interactions with stress

1. Little evolutionary change may occur at species borders since the cost of accommodating environmental stresses is high. Extreme examples of such stasis include cave animals in stable stressed environments and ‘living fossils’ in widely fluctuating stressed environments. 2. Variability from the molecular to the organismic level tends to be high under extreme stress. At the developmental level, the fitness of such variants may be low. This means that much developmental variability in natural populations may have little evolutionary significance. 3. Rapid evolutionary change of morphological traits is most likely to be based upon genes acting late in a developmental pathway under conditions which are ecologically and energetically permissive. 4. Although some increases in resistance to temperature extremes have been recorded in laboratory selection experiments, major extensions of extremes in natural populations appear difficult to achieve. The energetic costs of surviving extremes at species borders implies that the evolution of major developmental and morphological shifts is more likely to be a feature of populations of more equable habitats.