Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “REINFORCEMENT LEARNING”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

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

At least 1,351 records · Page 75Linked to original sources
PubMed1964

INTRACRANIAL REWARD DELAY AND THE ACQUISITION RATE OF A BRIGHTNESS DISCRIMINATION.

An application of the techniques of intracranial self-stimulation to the study of delayed reward indicates that the rate of discrimination learning for stimulation of the hypothalamus is a decreasing function of the delay interval. The resulting delay-of-reward gradient does not appear to differ appreciably from other such gradients based on food reward.

Discrimination Learning↗
PubMed2000

A characterization of approach and avoidance learning in high-alcohol-drinking (HAD) and low-alcohol-drinking (LAD) rats.

BACKGROUND: This study was undertaken as one of a series of experiments designed to examine basic behavioral characteristics present in rats bred specifically for alcohol preference. The basic premise for these experiments has been the idea that alcohol-preferring and -nonpreferring rats may differ in basic activation and inhibition control mechanisms that govern behavior and that different lines of alcohol-preferring rats may demonstrate differential deficits in behavioral activation and behavioral inhibition tendencies. In the present experiment, conditioned approach and avoidance behaviors were studied in alcohol-naïve high-alcohol-drinking (HAD), low-alcohol-drinking (LAD), and N/NIH rats to evaluate behavioral activation in this line of rats. METHODS: High alcohol drinking (HAD1), low alcohol drinking (LAD1), and N/Nih stock rats were trained to press a response bar during a tone signal to avoid a mild foot shock or receive a food reward. In addition, HAD2 and LAD2 rats, independently-bred replicate lines of the HAD1/LAD1 rats, were trained on the avoidance task. RESULTS: Although the HAD1 rats easily learned the appetitive version of the bar-pressing task, they did not learn the avoidance response. The LAD1 and N/Nih rats learned both the approach and the avoidance tasks normally. Similar to HAD1 rats, the HAD2 rats did not learn the avoidance response whereas LAD2 rats showed significant avoidance performance levels. CONCLUSIONS: The present data demonstrated that both HAD1 and HAD2 rats had a rather specific behavioral activation deficit: although they easily learned to press a bar to receive food reinforcement, they did not learn to press the bar to avoid a foot shock. We speculate that this failure to learn the avoidance response may be related to heightened anxiety in the HAD rats and that this excessive anxiety may lead to the development of high levels of alcohol consumption in these selectively bred rats.

Alcoholism↗
PubMed1996

Flavor-postingestive consequence associations incorporate the behaviorally opposing effects of positive reinforcement and anticipated satiety: implications for interpreting two-bottle tests.

Rats consuming distinctively flavored solutions containing different numbers of calories learn to associate each flavor with the ensuing postingestive effects ("flavor-postingestive consequence learning"). The hallmark of such learning is preferential consumption of one of the flavors in two-bottle tests (both flavors presented in nutrient-identical solutions). Two experiments were conducted to characterize the relationship between the number of calories associated with flavors and subsequent preferences in two-bottle tests. In Experiment 1, three groups of rats each associated distinctive flavors with real-fed sugar and with sham-fed sugar. The groups differed in the concentration of sucrose (8%, 14%, 24%) with which they were trained. In two-bottle tests rats trained with 8% sucrose preferred the real-fed flavor, whereas rats trained with 24% sucrose preferred the sham-fed flavor. Rats trained with 14% sucrose were intermediate to the other groups. In Experiment 2, two groups of rats associated distinctive flavors with two concentrations of real-fed sucrose. In two-bottle tests, the group trained with 1% and 5% sucrose preferred the flavor paired with the higher concentration, whereas the group trained with 5% and 30% sucrose preferred the flavor paired with the lower concentration. These findings suggest that flavor-postingestive consequence learning incorporates both positive reinforcement, which has the behavioral effect of increasing intake of the associated flavor, and anticipated satiety, which has the behavioral effect of suppressing intake of the associated flavor. When a flavor is associated with relatively few calories, the positive reinforcing effect predominates over anticipated satiety. However, when a flavor is associated with a greater number of calories, anticipated satiety predominates over the positive reinforcing effect.

Animals↗
PubMed2006

Olfactory learning in the one-day old rat: reinforcing effects of isoproterenol.

Within 24 h of their birth-induced norepinephrine surge, rat pups were tested for effects of a beta-receptor agonist, isoproterenol, on olfactory learning. Experiment 1 found no effect of isoproterenol on conditioning by pairing an odor (CS) with intraoral saccharin infusions. There was, however, unexpectedly strong responding in the unpaired control condition, which had the same contingency between the CS and isoproterenol as the paired condition. Experiment 2 found that pairings of odor and isoproterenol alone were sufficient for enhancing responding to the odor. Experiment 3 determined that isoproterenol had acted independently as a US for associative conditioning rather than facilitating nonassociative learning by mere exposure to the odor. These effects of isoproterenol as a US are consistent with the results of previous studies with older rats.

Adrenergic beta-Agonists↗
PubMed1998

Learning of sequential movements by neural network model with dopamine-like reinforcement signal.

Dopamine neurons appear to code an error in the prediction of reward. They are activated by unpredicted rewards, are not influenced by predicted rewards, and are depressed when a predicted reward is omitted. After conditioning, they respond to reward-predicting stimuli in a similar manner. With these characteristics, the dopamine response strongly resembles the predictive reinforcement teaching signal of neural network models implementing the temporal difference learning algorithm. This study explored a neural network model that used a reward-prediction error signal strongly resembling dopamine responses for learning movement sequences. A different stimulus was presented in each step of the sequence and required a different movement reaction, and reward occurred at the end of the correctly performed sequence. The dopamine-like predictive reinforcement signal efficiently allowed the model to learn long sequences. By contrast, learning with an unconditional reinforcement signal required synaptic eligibility traces of longer and biologically less-plausible durations for obtaining satisfactory performance. Thus, dopamine-like neuronal signals constitute excellent teaching signals for learning sequential behavior.

Algorithms↗
PubMed1959

Escape and avoidance conditioning in human subjects without their observation of the response.

An invisibly small thumb-twitch increased in rate of occurrence when it served, via electromyographic amplification, to terminate or postpone aversive noise stimulation. Subjects remained ignorant of their behavior and its effect. Their cumulative response curves resembled those obtained in similar work with animals. Other subjects, informed of the effective response, could not produce it deliberately in a size small enough to qualify for reinforcement.

Animals↗