Two experiments were run to demonstrate the presence of a partial reinforcement extinction effect (PREE) and a partial punishment effect (PPE) 4 weeks after training in a 1-trial/day procedure. In the PREE paradigm, two groups of animals were trained to run a straight alley for food reward; one group was rewarded on every trial (CRF), whereas the other was rewarded on only 50% of the trials (PRF). In the test phase, extinction, no reward was present (...) on any trial. Four weeks after the end of acquisition, and subsequent to 7 days of CRF retraining, PRF animals showed greater resistance to extinction than did their CRF controls. The PPE paradigm resembled the PREE paradigm in all aspects, except that mild electric foot shock, together with the food reward, was used instead of nonreward on 50% of the acquisition trials. In the test phase, all animals were shocked and given a food reward on every trial. Four weeks after the end of acquisition and retraining, the animals with previous experience of shock in acquisition showed greater resistance to punishment than did their controls. (shrink)

BACKGROUND -/- The importance of ontologies in the biomedical domain is generally recognized. However, their quality is often too poor for large-scale use in critical applications, at least partially due to insufficient training of ontology developers. -/- OBJECTIVE -/- To show the efficacy of guideline-based ontology development training on the performance of ontology developers. The hypothesis was that students who received training on top-level ontologies and design patterns perform better than those who only received training in (...) the basic principles of formal ontology engineering. -/- METHODS -/- A curriculum was implemented based on a guideline for ontology design. A randomized controlled trial on the efficacy of this curriculum was performed with 24 students from bioinformatics and related fields. After joint training on the fundamentals of ontology development the students were randomly allocated to two groups. During the intervention, each group received training on different topics in ontology development. In the assessment phase, all students were asked to solve modeling problems on topics taught differentially in the intervention phase. Primary outcome was the similarity of the students’ ontology artefacts compared with gold standard ontologies developed by the authors before the experiment; secondary outcome was the intra-group similarity of group members’ ontologies. -/- RESULTS -/- The experiment showed no significant effect of the guideline-based training on the performance of ontology developers (a) the ontologies developed after specific training were only slightly but not significantly closer to the gold standard ontologies than the ontologies developed without prior specific training; (b) although significant differences for certain ontologies were detected, the intra-group similarity was not consistently influenced in one direction by the differential training. -/- CONCLUSION -/- Methodologically limited, this study cannot be interpreted as a general failure of a guideline-based approach to ontology development. Further research is needed to increase insight into whether specific development guidelines and practices in ontology design are effective. (shrink)

Reinforcement learning approaches to cognitive modeling represent task acquisition as learning to choose the sequence of steps that accomplishes the task while maximizing a reward. However, an apparently unrecognized problem for modelers is choosing when, what, and how much to reward; that is, when (the moment: end of trial, subtask, or some other interval of task performance), what (the objective function: e.g., performance time or performance accuracy), and how much (the magnitude: with binary, categorical, or continuous values). In this (...) article, we explore the problem space of these three parameters in the context of a task whose completion entails some combination of 36 state–action pairs, where all intermediate states (i.e., after the initial state and prior to the end state) represent progressive but partial completion of the task. Different choices produce profoundly different learning paths and outcomes, with the strongest effect for moment. Unfortunately, there is little discussion in the literature of the effect of such choices. This absence is disappointing, as the choice of when, what, and how much needs to be made by a modeler for every learning model. (shrink)

The large scale, time varying, and diversification of physically coupled networked infrastructures such as power grid and transportation system lead to the complexity of their controller design, implementation, and expansion. For tackling these challenges, we suggest an online distributed reinforcement learning control algorithm with the one-layer neural network for each subsystem or called agents to adapt the variation of the networked infrastructures. Each controller includes a critic network and action network for approximating strategy utility function and desired control law, (...) respectively. For avoiding a large number of trials and improving the stability, the training of action network introduces supervised learning mechanisms into reduction of long-term cost. The stability of the control system with learning algorithm is analyzed; the upper bound of the tracking error and neural network weights are also estimated. The effectiveness of our proposed controller is illustrated in the simulation; the results indicate the stability under communication delay and disturbances as well. (shrink)

Obtaining prospective written consent from women to participate in trials when they are experiencing an obstetric emergency is challenging. Alternative consent pathways, such as gaining verbal consent at enrolment followed, later, by obtaining written consent, have been advocated by some clinicians and bioethicists but have received little empirical attention. We explored women’s and staff views about the consent procedures used during the internal pilot of a trial, where the protocol permitted staff to gain verbal consent at recruitment. Interviews with (...) staff and participating women. Data were analysed thematically and interviews were cross-compared to identify differences and similarities in participants’ views about the consent procedures used. Women and some staff highlighted benefits to obtaining verbal consent at trial enrolment, including expediting recruitment and reducing the burden on those left exhausted by their births. However, most staff with direct responsibility for taking consent expressed extreme reluctance to proceed with enrolment until they had obtained written consent, despite being comfortable using verbal procedures in their clinical practice. To account for this resistance, staff drew a strong distinction between research and clinical care and suggested that a higher level of consent was needed when recruiting into trials. In doing so, staff emphasised the need to engage women in reflexive decision-making and highlighted the role that completing the consent form could play in enabling and evidencing this process. While most staff cited their ethical responsibilities to women, they also voiced concerns that the absence of a signed consent form at recruitment could expose them to greater risk of litigation were an individual to experience a complication during the trial. Inexperience of recruiting into peripartum trials and limited availability of staff trained to take consent also reinforced preferences for obtaining written consent at recruitment. While alternative consent pathways have an important role to play in advancing emergency medicine research, and may be appreciated by potential recruits, they may give rise to unintended ethical and logistical challenges for staff. Staff would benefit from training and support to increase their confidence and willingness to recruit into trials using alternative consent pathways. This qualitative research was undertaken as part of the GOT-IT Trial. Date of registration 26/03/2014. (shrink)

We have analyzed many discrimination learning difficulties as reflecting multiple stimulus control topographies (SCTs). Nevin & Grace's analysis offers new variables to consider in the design of stimulus-control shaping procedures and cross-setting generalization of newly established behavior. A multiple-SCT perspective also suggests that fixed-trial discrimination procedures may offer advantages for reconciling momentum theory and partial reinforcement extinction effects.

This commentary considers factors that may account for the inconsistency between the behavioral momentum formulation and the partial reinforcement extinction effect. The method of testing, the variability of the schedule, the nature of the response-contingency, and response effort are considered. Some applications to real-world problems are also discussed.