Cross-participant modelling based on joint or disjoint feature selection: an fMRI conceptual decoding study

Multivariate classification techniques have proven to be powerful tools for distinguishing experimental conditions in single sessions of functional magnetic resonance imaging (fMRI) data. But they are vulnerable to a considerable penalty in classification accuracy when applied across sessions or participants, calling into question the degree to which fine-grained encodings are shared across subjects. Here, we introduce joint learning techniques, where feature selection is carried out using a held-out subset of a target dataset, before training a linear classifier on a source dataset. Single trials of functional MRI data from a covert property generation task are classified with regularized regression techniques to predict the semantic class of stimuli. With our selection techniques (joint ranking feature selection (JRFS) and disjoint feature selection (DJFS)), classification performance during cross-session prediction improved greatly, relative to feature selection on the source session data only. Compared with JRFS, DJFS showed significant improvements for cross-participant classification. And when using a groupwise training, DJFS approached the accuracies seen for prediction across different sessions from the same participant. Comparing several feature selection strategies, we found that a simple univariate ANOVA selection technique or a minimal searchlight (one voxel in size) is appropriate, compared with larger searchlights.

The welfare implications of large litter size in the domestic pig I:Biologica factors

Increasing litter size has long been a goal of pig breeders and producers, and may have implications for pig (Sus scrofa domesticus) welfare. This paper reviews the scientific evidence on biological factors affecting sow and piglet welfare in relation to large litter size. It is concluded that, in a number of ways, large litter size is a risk factor for decreased animal welfare in pig production. Increased litter size is associated with increased piglet mortality, which is likely to be associated with significant negative animal welfare impacts. In surviving piglets, many of the causes of mortality can also occur in non-lethal forms that cause suffering. Intense teat competition may increase the likelihood that some piglets do not gain adequate access to milk, causing starvation in the short term and possibly long-term detriments to health. Also, increased litter size leads to more piglets with low birth weight which is associated with a variety of negative long-term effects. Finally, increased production pressure placed on sows bearing large litters may produce health and welfare concerns for the sow. However, possible biological approaches to mitigating health and welfare issues associated with large litters are being implemented. An important mitigation strategy is genetic selection encompassing traits that promote piglet survival, vitality and growth. Sow nutrition and the minimisation of stress during gestation could also contribute to improving outcomes in terms of piglet welfare. Awareness of the possible negative welfare consequences of large litter size in pigs should lead to further active measures being taken to mitigate the mentioned effects. © 2013 Universities Federation for Animal Welfare.