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 1,279 records · Page 71Linked to original sources
PubMed2000

Neural circuits and topographic organization of the basal ganglia and related regions.

The present review was attempted to analyze the multiple channels of basal ganglia-thalamocortical connections, and the connections of their related nuclei. The prefrontal and motor areas consist of a number of modules, which seem to provide multiple subloops of the basal ganglia-thalamocortical connections in subhuman primates. There may be a great degree of convergence of the limbic, associative and motor loops at the level of the striatum, substantia nigra, pallidum, and the subthalamic nucleus as well as the pedunculopontine nucleus. Nigral dopaminergic neurons receive limbic input directly as well as indirectly through the striosomes in the striatum. Dopamine contributes to behavioral learning by signaling motivation and reinforcement. The pedunculopontine nucleus might be involved in behavioral state control, learning and reinforcement processes, locomotion and autonomic functions. Each subdivision of the motor areas receives a mixed and weighted transthalamic input from both the cerebellum and basal ganglia. In particular, based on the author's data, the hand/arm motor area and adjacent premotor area receive strong superficial basal ganglia-thalamocortical projections as well as the deep cerebello-thalamocortical projections. These areas, have very dense corticocotrical connections with other cortical areas, receive polymodal afferents from the parietal and temporal cortices, and integrated information, via multiple routes, from the prefrontal cortex. The author suggests that the ventrolateral part of the caudal medial pallidal segment (GPi) and the ventromedial part of the GPi are linked directly to these areas by ways of the oral part of ventral lateral nucleus (VLo) and the ventral part of the parvicellular part of ventral anterior nucleus (VApc), respectively. These connections are thought to be involved in the acquisition and coordination of motor sequences.

Animals↗
PubMed1963

General activity as instrumental: application to avoidance training.

Some data on Sidman avoidance in the pigeon are presented to illustrate a technique of instrumental training which places minimal restriction upon the form of the reinforced response. The effect of a warning signal is demonstrated, and the course of discrimination-reversal is described.

Animals↗
PubMed1963

Chaining and secondary reinforcement based on escape from shock.

Three white rats were trained to press a bar while being shocked. This produced a white noise. After 30 sec they were allowed to terminate both the shock and the noise by nosing a pigeon key. Comparison of the rates of pressing before and after the onset of the noise indicated that the noise itself was the immediate reinforcing agent for pressing. Furthermore, control tests showed that pressing was maintained only if it produced the noise: either omission of the noise or elimination of the dependency of the noise on the occurrence of the response led to a gradual abolition of pressing. When automatic termination of the shock was substituted for the key nosing requirement, however, only the key nosing extinguished. This indicated that the effectiveness of the noise as a reinforcer did not depend on its status as a discriminative stimulus for some other form of operant behavior.

Animals↗
PubMedDate not supplied

[Effect of acute and chronic deprivation of catecholaminergic system activity induced by 6-hydroxydopamine on the behavior of the rat].

On 202 male rats of Wistar line, a study was carried out of the effect of chronic and acute deprivations of the brain CA-systems activity resulting from administration of 6-OHDA on investigating behaviour and learning. Chronic deprivation of CA-systems activity by neonatal administration of 6-OHDA (100 mg/kg subcutaneously) and their acute deprivation by intracerebral administration of 6-OHDA to adult rats (150 mkg in each lateral ventriculus) was accompanied by similar deep changes of behaviour. Both forms of deprivation reduced the investigating activity of the animals in the open field. In both cases, the above 6-OHDA dozes sharply impeded the learning of animals with emotionally negative reinforcement, with no significant influence on learning with emotionally positive reinforcement. Both forms of deprivation of CA-systems activity weakened the reaction of frustration elicited by a sharp reduction of food reinforcement.

Animals↗
PubMed1986

Influence of isolation stress and inhibited protein biosynthesis on learning and memory in goldfish.

Goldfish, Carassius auratus auratus L. (Pisces, Cyprinidae), were trained by different kinds of training procedures under the influence of cycloheximide or puromycin, two inhibitors of the protein biosynthesis. After active avoidance training in a shuttle box an apparent amnesia was found only when the fish were exposed to a one day lasting isolation stress prior to training. If the animals were accustomed to isolation over a period of 20 days the inhibitors did not affect memory formation. After learning by positive reinforcement (food rewarded color discrimination) in groups under stress-free conditions, neither learning nor memory formation were impaired in spite of the presence of cycloheximide. It is suggested that the amnestic effect of the inhibitors is caused by isolation treatment. Lack of the additional stress, however, leads to memory formation.

Animals↗
PubMedDate not supplied

Effects of post-trial reinforcing vs. subreinforcing stimulation of the substantia nigra on passive avoidance learning.

Experiments were designed to investigate the role of post-trial reinforcing and subreinforcing stimulation of the substantia nigra on memory processing. Thirty sec post-trial reinforcing stimulation (0.2 sec on/1.8 sec off) impaired learning of response suppression in a step-down task compared to control animals as well as to animals stimulated at subthreshold current level (i.e., at 25% of the current level shown to maintain optimal self-stimulation in previously performed self-stimulation sessions). In a second small-box passive avoidance experiment, i.e., the alcove-avoidance task, opposite results were attained: Subreinforcing stimulation attenuated learning whereas neither suprathreshold stimulated animals nor control animals showed impairment of learning. The conclusions drawn from these results are as follows: Post-trial stimulation of the substantia nigra interferes with memory processing. This attenuation of learning is obviously task-dependent and can additionally be influenced by the quality of the stimulation, i.e., whether it is reinforcing or not. Possible explanations to account for these task-dependent and quality of stimulation-dependent effects of post-trial substantia nigra stimulation are discussed.

Animals↗
PubMed1959

Discrimination learning.

Eight rats were run through discrimination training sessions in which responses in the dark were not reinforced whereas the first response after the onset of a light was reinforced. The procedure generated orderly learning and latency data for the individual animal. The latency distributions are adequately described by a simple mathematical formulation.

Animals↗
PubMed1976

Control of fingertip temperature increases via biofeedback in learning-disabled and normal children.

Although considerable research has been done with biofeedback in adults, little is known of its effect in children of different ages or those with learning disabilities. This study assessed the effects of thermal biofeedback in 60 children (40 boys, 20 girls) aged 7 to 9 yr. (Mage 8-6) half of whom were learning-disabled and half, normal children matched for age, sex, grade, race, socioeconomic status, and IQ. Training consisted on one 15-min. period daily for five days with three trials per period. Children (16 learning-disabled, 16 normals) received consistent reinforcement for digital temperature increases while 28 children (14 learning-disabled, 14 normals) received mixed reinforcement after Day 1, on which all children were consistently reinforced to yield a performance baseline. Feedback was provided by a variable intensity light and toy electric train. Learning was demonstrated only for the consistently reinforced group, which performed almost twice as well as those receiving mixed reinforcement. Learning-disabled children learned thermal control even better than normals, explained in terms of biofeedback reinforcing an internal steady state conducive to learning. Younger children did better than older children, and girls did somewhat better than boys. Post-training improvement in figure-ground discrimination and intersensory-integration was linked with performance and learning, respectively.

Adult↗
PubMed1980

The effect of reinforcement and verbal rehearsal on selective attention in learning-disabled children.

A central-incidental task of selective attention was administered to 100 learning-disabled boys--48 younger children (8 1/2-10 1/2 years) and 52 older children (10 1/2-12 1/2 years). Subjects at both age levels were assigned to one of four conditions: (a) a standard condition; (b) a rehearsal condition, in which subjects were taught a verbal rehearsal strategy; (c) a reinforcement condition, in which correct responses were rewarded; and (d) a combined rehearsal-reinforcement condition. Older subjects recalled more central task but not more incidental task information than younger subjects. A measure of selective attention efficiency was also greater for older than for younger subjects. Central recall in the rehearsal-reinforcement condition was greater than in any other condition. Central recall was greater in the rehearsal condition than in the standard and reinforcement conditions. Selective attention efficiency was greater for both rehearsal conditions than for both nonrehearsal conditions. Incidental recall was higher for reinforcement subjects than for rehearsal subjects. Results suggest that induced verbal rehearsal improves central recall and selective attention in learning-disabled children. Reinforcement alone does not improve central recall but may when paired with rehearsal.

Attention↗
PubMed1993

Cholecystokinin-dopamine interactions within the nucleus accumbens in the control over behaviour by conditioned reinforcement.

Cholecystokinin (CCK) is colocalised with dopamine in the postero-medial nucleus accumbens (NAS). We have utilised an acquisition of a new response procedure to investigate the interaction between CCK and dopamine in the control over behaviour by conditioned reinforcers. A conditioned reinforcer (CR) may be defined as an initially neutral stimulus which gains control over behaviour through selective association with a primary reinforcer. Here, rats learned to associate a light/noise compound stimulus with the imminent availability of 10% sucrose reinforcement. Later, in the absence of sucrose, responding on one of two novel levers (the CR lever) was acquired and maintained by contingent presentation of the CR alone, while responding on the second lever had no programmed consequences. In Expt. 1, infusion of 10 micrograms D-amphetamine within the postero-medial NAS enhanced responding selectively on the CR lever. Infusion of sulphated CCK octapeptide (CCK: 1 or 10 ng) alone within the same area had no effect on response rate. However, infusion of CCK immediately prior to D-amphetamine caused a dose-dependent potentiation of the impact of D-amphetamine upon rates of response on the CR lever. In Expt. 2, infusion of D-amphetamine (10 micrograms) within the postero-medial NAS again enhanced responding selectively upon the CR lever. Intra-accumbens infusion of CCK (10 ng), or s.c. administration of the CCKA receptor antagonist devazepide had no effect upon response rates. However, CCK again potentiated the D-amphetamine-induced increase in rates of response, and this potentiation was blocked by pretreatment with devazepide. These results are discussed in terms of the co-modulation by CCK and dopamine of the processing of reward-related stimuli within the NAS.

Afferent Pathways↗
PubMed1984

Gangliosides minimize behavioral deficits and enhance structural repair after brain injury.

Injections of GM1-gangliosides (30 mg/kg, i.p.) in adult rats were shown to reduce behavioral deficits after brain lesions. This was observed (1) after bilateral electrolytic lesions of the caudate nucleus in a learning task involving negative reinforcement; (2) following aspiration lesions of the mediofrontal cortex in a learning task involving positive reinforcement; and (3) when rotational behavior was assessed after amphetamine or apomorphine injections in animals with partial hemitransections of the nigro-striato-nigral fibers. A detailed anatomical analysis of the latter study, using a retrograde tract-tracing dye wheat germ agglutinin-horseradish peroxidase (WGA-HRP), provided evidence for ganglioside-stimulated, neuronal reorganization of connections to the caudate nucleus. Our findings support the notion that gangliosides reduce behavioral deficits following brain injury by preventing secondary neuronal degeneration and/or enhancing structural reorganization of remaining afferents, rather than by influencing denervation supersensitivity.

Animals↗