Spatio-temporal credit assignment in population learning

We present a model for plasticity induction in reinforcement learning which is based on a cascade of synaptic memory traces. In the cascade of these so called eligibility traces presynaptic input is first corre lated with postsynaptic events, next with the behavioral decisions and finally with the external reinforcement. A population of leaky integrate and fire neurons endowed with this plasticity scheme is studied by simulation on different tasks. For operant co nditioning with delayed reinforcement, learning succeeds even when the delay is so large that the delivered reward reflects the appropriateness, not of the immediately preceeding response, but of a decision made earlier on in the stimulus - decision sequence . So the proposed model does not rely on the temporal contiguity between decision and pertinent reward and thus provides a viable means of addressing the temporal credit assignment problem. In the same task, learning speeds up with increasing population si ze, showing that the plasticity cascade simultaneously addresses the spatial problem of assigning credit to the different population neurons. Simulations on other task such as sequential decision making serve to highlight the robustness of the proposed sch eme and, further, contrast its performance to that of temporal difference based approaches to reinforcement learning.

Patches of Bare Ground as a Staple Commodity for Declining Ground-Foraging Insectivorous Farmland Birds

Conceived to combat widescale biodiversity erosion in farmland, agri-environment schemes have largely failed to deliver their promises despite massive financial support. While several common species have shown to react positively to existing measures, rare species have continued to decline in most European countries. Of particular concern is the status of insectivorous farmland birds that forage on the ground. We modelled the foraging habitat preferences of four declining insectivorous bird species (hoopoe, wryneck, woodlark, common redstart) inhabiting fruit tree plantations, orchards and vineyards. All species preferred foraging in habitat mosaics consisting of patches of grass and bare ground, with an optimal, species-specific bare ground coverage of 30–70% at the foraging patch scale. In the study areas, birds thrived in intensively cultivated farmland where such ground vegetation mosaics existed. Not promoted by conventional agri-environment schemes until now, patches of bare ground should be implemented throughout grassland in order to prevent further decline of insectivorous farmland birds.

Spatio-Temporal Credit Assignment in Neuronal Population Learning

n learning from trial and error, animals need to relate behavioral decisions to environmental reinforcement even though it may be difficult to assign credit to a particular decision when outcomes are uncertain or subject to delays. When considering the biophysical basis of learning, the credit-assignment problem is compounded because the behavioral decisions themselves result from the spatio-temporal aggregation of many synaptic releases. We present a model of plasticity induction for reinforcement learning in a population of leaky integrate and fire neurons which is based on a cascade of synaptic memory traces. Each synaptic cascade correlates presynaptic input first with postsynaptic events, next with the behavioral decisions and finally with external reinforcement. For operant conditioning, learning succeeds even when reinforcement is delivered with a delay so large that temporal contiguity between decision and pertinent reward is lost due to intervening decisions which are themselves subject to delayed reinforcement. This shows that the model provides a viable mechanism for temporal credit assignment. Further, learning speeds up with increasing population size, so the plasticity cascade simultaneously addresses the spatial problem of assigning credit to synapses in different population neurons. Simulations on other tasks, such as sequential decision making, serve to contrast the performance of the proposed scheme to that of temporal difference-based learning. We argue that, due to their comparative robustness, synaptic plasticity cascades are attractive basic models of reinforcement learning in the brain.

Potential and limitations of organic and fair trade cotton for improving livelihoods of smallholders: evidence from Central Asia

Cotton is a leading agricultural non-food commodity associated with soil degradation, water pollution and pesticide poisoning due to high levels of agrochemical inputs. Organic farming is often promoted as a means of addressing the economic, environmental and health risks of conventional cotton production, and it is slowly gaining ground in the global cotton market. Organic and fair trade cotton are widely seen as opportunities for smallholder farmers to improve their livelihoods thanks to higher returns, lower input costs and fewer risks. Despite an increasing number of studies comparing the profitability of organic and non-organic farming systems in developing and industrialized countries, little has been published on organic farming in Central Asia. The aim of this article is to describe the economic performance and perceived social and environmental impacts of organic cotton in southern Kyrgyzstan, drawing on a comparative field study conducted by the author in 2009. In addition to economic and environmental aspects, the study investigated farmers’ motivations toward and assessment of conversion to organic farming. Cotton yields on organic farms were found to be 10% lower, while input costs per unit were 42% lower; as a result, organic farmers’ cotton revenues were 20% higher. Due to lower input costs as well as organic and fair trade price premiums, the average gross margin from organic cotton was 27% higher. In addition to direct economic benefits, organic farmers enjoy other benefits, such as easy access to credit on favorable terms, provision of uncontaminated cottonseed cooking oil and cottonseed cake as animal feed, and marketing support as well as extension and training services provided by newly established organic service providers. The majority of organic farmers perceive improved soil quality, improved health conditions, and positively assess their initial decision to convert to organic farming. The major disadvantage of organic farming is the high manual labor input required. In the study area, where manual farm work is mainly women's work and male labor migration is widespread, women are most affected by this negative aspect of organic farming. Altogether, the results suggest that, despite the inconvenience of a higher workload, the advantages of organic farming outweigh its disadvantages and that conversion to organic farming improves the livelihoods of small-scale farmers.

“All Data is Credit Data,” or, On Close Reading as a Reciprocal Process in Digital Knowledge Environments

The new knowledge environments of the digital age are oen described as places where we are all closely read, with our buying habits, location, and identities available to advertisers, online merchants, the government, and others through our use of the Internet. This is represented as a loss of privacy in which these entities learn about our activities and desires, using means that were unavailable in the pre-digital era. This article argues that the reciprocal nature of digital networks means 1) that the privacy issues that we face online are not radically different from those of the pre-Internet era, and 2) that we need to reconceive of close reading as an activity of which both humans and computer algorithms are capable.