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 1,117 records · Page 62Linked to original sources
PubMed2005

More evidence of educational inadequacies in musculoskeletal medicine.

In their study, Freedman and Bernstein suggested that 80% of a group of graduates from many of the best medical schools in the United States were deficient in their knowledge of basic facts and concepts in musculoskeletal medicine. How do these results compare with results from students attending a medical school with a long-standing dedicated program to musculoskeletal education? Does additional clinical experience in musculoskeletal medicine improve understanding of the basic facts and concepts introduced in a second-year course? A modified version of an exam used to assess the competency of incoming interns at the University of Pennsylvania was used to assess the competency of medical students during various stages of their training at the University of Washington. Despite generally improved levels of competency with each year at medical school, less than 50% of fourth-year students showed competency. Students who completed a musculoskeletal clinical elective scored higher and were more competent (78%) than students who did not take an elective. These results suggested that the curricular approach toward teaching musculoskeletal medicine at this medical school was insufficient and that competency increased when learning was reinforced during the clinical years.

Clinical Competence↗
PubMed2004

Adaptive probabilistic neural networks for pattern classification in time-varying environment.

In this paper, we propose a new class of probabilistic neural networks (PNNs) working in nonstationary environment. The novelty is summarized as follows: 1) We formulate the problem of pattern classification in nonstationary environment as the prediction problem and design a probabilistic neural network to classify patterns having time-varying probability distributions. We note that the problem of pattern classification in the nonstationary case is closely connected with the problem of prediction because on the basis of a learning sequence of the length n, a pattern in the moment n + k, k > or = 1 should be classified. 2) We present, for the first time in literature, definitions of optimality of PNNs in time-varying environment. Moreover, we prove that our PNNs asymptotically approach the Bayes-optimal (time-varying) decision surface. 3) We investigate the speed of convergence of constructed PNNs. 4) We design in detail PNNs based on Parzen kernels and multivariate Hermite series.

Algorithms↗
PubMed1998

A perspective on the control of FES-supported standing.

This special section is about the control of electrical stimulators to restore standing functions to paraplegics. It addresses several important topics regarding the interactions of the intact central nervous systems (CNS) with the artificial control system. The topics are as follows: how paraplegics use their arms to help themselves stand up with functional electrical stimulation (FES); the user-driven artificial control of FES-supported standing up; a controller which is promising for the control of sitting down; the application of reinforcement machine learning for the controllers of standing up; arms-free standing with voluntary upper body balancing and artificially controlled ankle stiffness; and cognitive feedback in balancing. This Commentary introduces the papers in this section and relates them to earlier research.

Electric Stimulation Therapy↗
PubMed2006

Developing and evaluating an interactive information skills tutorial.

OBJECTIVE: To develop and evaluate a web-based interactive information skills tutorial integrated into the curriculum. To determine whether the tutorial was acceptable to students and explore the use of a skills assessment tool in identifying whether the tutorial improved skills. METHODS: The development of a tutorial on OVID medline to teach transferable information skills. A small cohort study to evaluate students' views on the tutorial and its effects on information skills. RESULTS: Thirteen objective assessments were usable. There was a statistically significant improvement in mean final assessment scores, compared with mean pre-training scores, F(2,14) = 11.493, P = 0.001. Eleven (85%) students had improved their overall information skills. The improvement in overall searching skills was enhanced by referral to the tutorial. CONCLUSIONS: The tutorial was successfully developed and integrated into a Masters programme curriculum. In this setting, it appears to reinforce active learning, and was well received by students, who developed core generic searching skills and demonstrated improved information skills in the short and longer term. Students could use the tutorial for revision and study at a time and place of their choosing. Further evaluation is required to assess the impact of using the tutorial with large groups of students, and as a stand-alone teaching medium.

Adult↗
PubMed1963

Prediction of the comparative reinforcement values of running and drinking.

The probability of free drinking and running in rats was controlled by sucrose concentration and force requirements on an activity wheel. Drinking and running were then made contingent on pressing a bar. Barpressing increased monotonically with the associated response probability, and equally for drinking and running. The results support the assumption that different responses of equal probability have equal reinforcement value.

Alcohol Drinking↗
PubMed1963

Intracranial self-stimulation in man.

Intracranial self-stimulation techniques, modified for use with neuropsychiatric patients, provide data suggestive of positive and negative reinforcing properties of brief electrical stimulation to various subcortical structures of the human brain.

Brain↗
PubMed1984

Operant conditioning and the modulation of cardiovascular function.

Operant learning appears to be one of the primary mechanisms underlying what cardiovascular and pulmonary physiologists have called adaptation, habituation, and central command. In general, studies that have attempted to use operant conditioning alone to create experimental models of behaviorally induced disease have been unsuccessful because the cardiovascular responses adapted or habituated over time. Thus, these studies have provided implicit demonstrations of the roles played by CNS and conditioning processes in achieving and preserving homeostasis. During the past few years those interested in behavioral contributions to cardiovascular pathology have therefore begun to look at interactions between behavior and other variables that might predispose organisms towards pathology (e.g. genetic background; excessive sodium intake). Perhaps even more promising has been the growth in the number of technically competent, well-controlled studies designed to investigate: (a) broad scientific questions of how behaviorally important processes such as learning and reinforcement interact with physiologically important variables such as blood flow redistribution and cardio-pulmonary integration; and (b) the role of behavioral variables in CNS control of the circulation. Based upon our survey of the recent literature, we believe that the time is ripe for those interested in cardiovascular neurobiology increasingly to include behavioral variables in their studies, because the raison d'etre of the CNS is to optimize the organism's ability to interact with its environment. Only when these organismic-environmental interactions are studied both behaviorally and physiologically, in a broad biological context, will it be possible to develop rational models of neuro-circulatory regulation.

Animals↗
PubMed2002

Integrating recent advances in neuroscience into undergraduate neuroscience and physiology courses.

Neuroscience has enjoyed tremendous growth over the past 20 years, including a substantial increase in the number of neuroscience departments, programs, and courses at the undergraduate level. To meet the need of new neuroscience courses, there has also been growth in the number of introductory neuroscience textbooks designed for undergraduates. However, textbooks typically trail current knowledge by five to ten years, especially in neuroscience where our understanding is increasing rapidly. Consequently, it is often important to supplement neuroscience and physiology textbooks with information about recent findings in neuroscience. To design supplementary educational material, it is essential first to identify the educational objectives of the program and the characteristics of the learners, which can differ dramatically between undergraduate and graduate or professional students. Four principles that may serve the selection and design of supplementary material for undergraduate neuroscience and physiology courses are that (1) material must be interesting to the undergraduates, (2) material should reinforce previously learned concepts, (3) students must be adequately prepared, and (4) the teacher and student must have sufficient appropriate resources.

Animals↗
PubMed1976

Designing environments for mentally retarded clients.

An environment that stimulates the auditory, visual, and tactile senses increases appropriate behavior among mentally retarded clients and thereby reinforces their learning activities. The author describes several low-cost deisgn techniques for giving a more stimulating and personalized atmosphere to large, existing institutions; discusses design requirements for group homes for retarded and handicapped clients in the community; and summarizes design ideas to be incorporated in new facilities built for retarded clients. She describes the implementation of her design ideas on a unit for blind retarded clients at the Mansfield Training School in Connecticut.

Community Health Services↗
PubMed2002

Serial killers with military experience: applying learning theory to serial murder.

Scholars have endeavored to study the motivation and causality behind serial murder by researching biological, psychological, and sociological variables. Some of these studies have provided support for the relationship between these variables and serial murder. However, the study of serial murder continues to be an exploratory rather than explanatory research topic. This article examines the possible link between serial killers and military service. Citing previous research using social learning theory for the study of murder, this article explores how potential serial killers learn to reinforce violence, aggression, and murder in military boot camps. As with other variables considered in serial killer research, military experience alone cannot account for all cases of serial murder. Future research should continue to examine this possible link.

Homicide↗
PubMed2006

Does attentional cueing affect dichotic listening performance in children?

This study addresses the effects of verbal versus nonverbal (tone) shifts of attention on dichotic listening (DL) performance with children. Theoretically, a tonal cue may be more effective in increasing attention than a verbal cue following instruction. The inconsistency of studies reporting substantial effects of attention on ear asymmetries in children with or without learning disabilities (LDs) may be due to a developmental difference in their ability to use verbal or tone cues to select stimuli for recall. Participants included 30 right-handed children (15 control, 15 with LDs) with a mean age of 10.8 years. Each participant received 60 trials of a monaural tone cue task, 60 trials of a binaural verbal cue task, and 60 trials of a monaural verbal cue task, to direct attention to either the left or right ear before the presentation of consonant-vowel syllable pairs in a DL task. A factorial design analysis of variance yielded a significant right-ear advantage for both groups. More important, the Group x Task interaction was found to be significant, indicating that group performance on ear scores was dependent on type of cueing condition. Whereas all 3 cue conditions were effective in orienting attention for control participants, larger shifts were apparent under both binaural and monaural verbal instructional cue conditions. In contrast, participants with LD showed larger shifts of attention under the tonal cue condition. These results show that control participants have greater ability to focus attention with the use of a verbal cue, whereas participants with LD show greater ability to orient attention with the use of a tone cue in reducing error rates in DL performance.

Acoustic Stimulation↗
PubMed2006

The utilization of systematic outcome mapping to improve performance management in health care.

Performance management is an important mechanism for ensuring accountability and improving the quality of health-care services. The last decade has witnessed a proliferation in the development of performance measurement systems for assessing health-care processes and outcomes at the program, hospital, district, system and national level. This has allowed for comparison and benchmarking between similar aspects of care at each of these levels. Unfortunately, most performance systems are devoid of clear mechanisms for translating feedback from measures into strategies for action, thus leaving largely unfulfilled the quality and management aspect necessary to improve health-care services. Therefore, the thinking that goes into designing these systems must change. This article outlines a management framework called systematic outcome mapping that provides for performance management rather than just performance measurement by allowing for quality improvement to be built into performance indicator development. It utilizes evidence-based medicine and expert consensus opinion to establish linkages between processes of care and their outcomes with the clear intent that feedback from information provided by performance indicators can be used to modify health-care activities so as to improve health outcomes. This fulfils the quality improvement aspect of performance measurement and makes it an integral part of a performance management framework that reinforces organizational learning through feedback from outcomes and the assessment of organizational routines.

Delivery of Health Care↗
PubMed1978

Behavioral assessment of visual toxicity.

A wide variety of behavioral methods has been employed with animals to assess visual changes induced by drugs or toxicants. The methods range from simple to complex, from broad screening devices to narrowly focused techniques. Their relative advantages for the environmental toxicologist are discussed. Manipulation of stimulus values is an essential ingredient in the identification of specific sensory functions. The percentage of correct choices from a discrete-trial, multiple-choice discrimination procedure is to be preferred to measures of response rate, speed or reaction time when experiments require answers about specific visual functions.

Animals↗