PubMed Health⌕ Search

SEARCH · PubMed Health

Results for “Reinforcement learning”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,207 records · Page 67Linked to original sources
PubMed1975

Escalation of aggression: experimental studies.

A finding commonly obtained in research using the Buss "aggression machine" is a main effect for trail blocks, indicating an escalation in shock intensity over trails. Theoretical explanations for this effect were tested in a modified verbal operant-conditioning situation. In Experiment 1, subjects could administer any of 10 levels of positive reinforcement to a "learner" for correct verbal responses or any of 10 levels of negative reinforcement to a learner for incorrect responses. Half of the subjects were required to begin with weak, half with strong, reinforcements. Results indicated that, regardless of condition, subjects gave more intense reinforcements as the learning trails progressed. Those who administered negative reinforcements devalued the learner relative to those who administered positive reinforcements. In Experiment 2, a role-playing procedure was used in which subjects administered either positive or negative reinforcements to a learner whose performance either did or did not improve over trials. Again, in all experimental groups, subjects administered increasingly intense reinforcements over trials. The results are interpreted as supporting a disinhibition theory of anti- and prosocial behavior.

Aggression↗
PubMed1992

Early onset of age-related impairment of aversive and appetitive learning in the SAM-P/8 mouse.

SAM-P/8 (P/8if) mice, trained on aversive and appetitive training tasks, showed age-related impairment of learning on more difficult tasks. Using aversive training tasks, acquisition of step-down passive avoidance did not differ between 4- and 12-month-old mice. T-maze footshock avoidance acquisition detected impairment from 9 to 12 months of age. On the appetitive tasks, 12-month-old mice required significantly more trails than 4-month-old mice to learn to press a lever for milk reinforcement or to learn a specific sequence of 10 black-white, left-right discriminations. However, learning a simple black-white discrimination in the same 10-unit maze did not discriminate between the performance of 4- and 12-month-old mice. A lack of age-related changes in food, water, and milk consumption suggests that changes in motivation do not account for the decline in learning in older P/8if mice. The results suggest an early onset of impaired learning relative to the 17.2-month median life span of the P/8if mice.

Aging↗
PubMed1977

Effects of social reinforcement for visual attention on classroom learning by disadvantaged preschoolers.

To investigate the relationship between visual attention and classroom learning in a disadvantaged preschool population students' attending was experimentally maintained at high levels, using a social reinforcement procedure, and recall of visual and auditory educational materials was examined. Subjects were 16 white children of low income whose mean age was 5.1 yr. and who were students at a Head Start center. A two-group, randomized-blocks design was employed. Reinforcement of attending resulted in greater recall of auditory stimuli by experimental subjects, while a decline was found in recall performance by control subjects. Recall of visual material was not significantly affected by the treatment. Correlations between attending and both auditory and visual recall were statistically significant.

Attention↗
PubMed1963

Fixed-ratio escape reinforcement.

Escape responses of squirrel monkeys were reinforced according to a fixed-ratio schedule. The reinforcement was a period of safety from a stimulus that signalled the delivery of intermittent pain-shocks. When the frequency of shock was gradually reduced, the performance remained at a high level until the shocks were quite infrequent. Similarly, the duration of the period of safety could be reduced to a few seconds with little loss of behavior. Thus, the responses appeared to be reinforced by even a brief period of safety, the actual degree of shock reduction being fairly slight. The changes in responding during this fixed-ratio escape procedure were comparable to the response changes typically obtained during fixed-ratio food procedures.

Animals↗
PubMed1963

Stimulus generalization of a positive conditioned reinforcer.

Stimulus generalization has been observed for discriminative, eliciting, and emotional functions of stimuli. In our study, in order to investigate the generalization of the reinforcing function of stimuli, pigeons were trained in a Skinner box to peck at an unlighted key to obtain aperiodic, brief exposures of light at a wavelength of 550 mmicro, the positive conditioned reinforcer, which was immediately followed by food reward. Testing in extinction, we obtained generalization gradients for the number of responses and the time the pigeons expended to produce exposures on the unlighted key of 550 mmicro, 530 mmicro) 510 mmicro, or no light. This finding suggests that stimulus generalization occurs with all functions of stimuli.

Animals↗
PubMed1994

Computer-assisted shock-reinforced Y-maze training: a method for studying spatial alternation behaviour.

Y-mazes are widely used for discrimination learning and spatial alternation tasks. We present here a computer-assisted Y-maze model for spatial alternation learning via footshock reinforcement. No intramaze cues are provided and handling between trials is no longer required. Pretraining application of scopolamine (1 mg kg-1) and D-AP5 (2-amino-5-phosphonopentanoate; 0.016 mg) clearly impairs acquisition and retention (tested 24 h after acquisition) of spatial alternation. These effects are not due to state-dependent changes caused by drug treatment. Similar results were reported by others in food-motivated spatial alternation tasks as well as spontaneous alternation paradigms. The data support our model as a useful tool for studying spatial alternation in the Y-maze.

2-Amino-5-phosphonovalerate↗
PubMed2003

Discrimination between outcomes in instrumental learning: effects of preexposure to the reinforcers.

In two experiments rats received instrumental training with two response levers, one response being reinforced by sucrose solution and the other by sucrose pellets. Prior to a test session, on which both levers were made available in the absence of reinforcement, the rats were given free access to one of the reinforcers, a procedure known to reduce its value. It was found that the rats responded at a lower rate on the lever that had produced the now-devalued reinforcer, but that this effect was substantial only in rats that had received preexposure to the two reinforcers before instrumental training was begun (Experiment 1). Experiment 2 demonstrated that this effect was obtained only when presentations of the two reinforcers were presented according to an inter-mixed schedule during preexposure. It is suggested that this result constitutes an instance of the perceptual learning effect in which intermixed preexposure to similar events enhances their discriminability.

Animals↗
PubMedDate not supplied

Learning by neurones: role of attention, reinforcement and behaviour.

The importance of the behavioural situation, attentional demands of the task, and stimulus-reinforcement contingencies in promoting or permitting experience-dependent neuronal plasticity is argued. Evidence is provided for the specific activation of the locus coeruleus noradrenergic system of the rat by novel stimuli encountered while investigating the environment, as well as during a formal learning situation. Noradrenergic neurons are particularly concerned with changes in the predictive value of the stimulus, when new learning should occur. Noradrenaline, released at LC terminals in target sensory systems, could facilitate shifts in attention, information processing and memory through its well-documented gating and tuning effects and its permissive role in long-term potentiation. Dopamine neurons, which fire persistently to reward during learning, could be involved in maintaining the behavioural response.

Animals↗
PubMed2002

Psychological myths in e-learning.

Traditional education and training has paid scant attention to the psychology of learning. Despite detailed research into motivation, distribution and reinforcement, most current methods of delivery still rely on a supply-led, lecture and classroom-based model that flies in the face of the theory. With e-learning we have a chance to reflect on this gap between theory and practice. E-learning, in the sense of web-based learning, is a new discipline but the psychology of learning has a much longer pedigree. This paper relates some common myths about e-learning back to some major themes in the psychology of learning. Is e-learning faster and more effective? Many people get the wrong learning at the wrong time. Can e-learning help with prerequisite knowledge? Should the learning be massed or distributed, i.e. all at once or little and often? There are also the issues of motivation and cognitive engagement. How can e-learning motivate learners or how can we motivate learners into using this new medium? What type of cognitive engagement is necessary for learning? Traditional 'sheep-dip' methods of learning are poor on reinforcement. Can e-learning help reinforce learning?

Education, Distance↗