SEARCH · PubMed Health

Results for “Predictive Learning Models”

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

PubMedDate not supplied

Age-related impairment in the 250-millisecond delay eyeblink classical conditioning procedure in C57BL/6 mice.

In this study we tested 4-, 9-, 12-, and 18-month-old C57BL/6 mice in the 250-msec delay eyeblink classical conditioning procedure to study age-related changes in a form of associative learning. The short life expectancy of mice, complete knowledge about the mouse genome, and the availability of transgenic and knock-out mouse models of age-related impairments make the mouse an excellent species for expanding knowledge on the neurobiologically and behaviorally well-characterized eyeblink classical conditioning paradigm. Based on previous research with delay eyeblink conditioning in rabbits and humans, we predicted that mice would be impaired on this cerebellar-dependent associative learning task in middle-age, at ~9 months. To fully examine age differences in behavior in mice, we used a battery of additional behavioral measures with which to compare young and older mice. These behaviors included the acoustic startle response, prepulse inhibition, rotorod, and the Morris water maze. Mice began to show impairment in cerebellar-dependent tasks such as rotorod and eyeblink conditioning at 9 to 12 months of age. Performance in hippocampally dependent tasks was not impaired in any group, including 18-month-old mice. These results in mice support results in other species, indicating that cerebellar-dependent tasks show age-related deficits earlier in adulthood than do hippocampally dependent tasks.

Acoustic Stimulation↗

PubMedDate not supplied

The applicability of the theory of planned behavior to the intention to quit smoking across workplaces in southern Taiwan.

An examination of the applicability of the theory of planned behavior (TPB) to the intention to quit smoking across workplaces was conducted. Subjects were randomly selected from three workplaces in southern Taiwan. Those from a large public steel-manufacturing company were used for model building, and those from two private auto-parts-manufacturing companies served to cross-validate the model. Eligible subjects were divided into three study samples: a learning sample and two test samples. Three predictors--priority of quitting, past behavior (measured as previous quit attempt), and habit (measured as nicotine dependence)--were added to the TPB model. The results of this study show that TPB based on the learning sample fit well in another sample from the same workplace but poorly in other workplaces. When priority of quitting and past behavior were added to the TPB model, prediction to other workplaces significantly improved. Habit had no significant contribution to the intention to quit in the TPB model. Detailed discussions of the results are provided.

Adult↗

PubMed1987

Intellectual functioning and aggression.

In a 22-year study, data were collected on aggressiveness and intellectual functioning in more than 600 subjects, their parents, and their children. Both aggression and intellectual functioning are reasonably stable in a subject's lifetime and perpetuate themselves across generations and within marriage pairs. Aggression in childhood was shown to interfere with the development of intellectual functioning and to be predictive of poorer intellectual achievement as an adult. Early IQ was related to early subject aggression but did not predict changes in aggression after age 8. On the other hand, differences between early IQ and intellectual achievement in middle adulthood were predictable from early aggressive behavior. A dual-process model was offered to explain the relation between intellectual functioning and aggressive behavior. We hypothesized that low intelligence makes the learning of aggressive responses more likely at an early age, and this aggressive behavior makes continued intellectual development more difficult.

Achievement↗

PubMed1995

Interactions of stimulus attributes, base rates, and feedback in recognition.

Continuous old-new recognition was studied in relation to 3 factors that have been relatively neglected in previous research-stimulus attributes, old-new base rates, and informative feedback following responses. Under all conditions, both hits and false alarms increased over trials and all measures of recognition depended strongly on stimulus properties, notably interitem similarity. In contrast to expectations based on earlier results, both hit and false-alarm levels proved independent of old-new base rate when tests were given without feedback; with feedback added, false-alarm rates tended to approach true old-stimulus base rates with some types of stimuli, though not with words. The findings are compatible, in general, with current composite-memory models and were predicted in detail by an array-similarity model deriving from categorization theory.

Adult↗

PubMed2003

An application of signal detection theory with finite mixture distributions to source discrimination.

A mixture extension of signal detection theory is applied to source discrimination. The basic idea of the approach is that only a portion of the sources (say A or B) of items to be discriminated is encoded or attended to during the study period. As a result, in addition to 2 underlying probability distributions associated with the 2 sources, there is a 3rd distribution that represents items for which sources were not attended to. Thus, over trials, the observed response results from a mixture of an attended (A or B) distribution and a nonattended distribution. The situation differs in an interesting way from detection in that, for detection, there is mixing only on signal trials and not on noise trials, whereas for discrimination, there is mixing on both A and B trials. Predictions of the mixture model are examined for data from several recent studies and in a new experiment.

Association Learning↗

PubMed2001

Forming classes by stimulus frequency: behavior and theory.

Visual classification is the way we relate to different images in our environment as if they were the same, while relating differently to other collections of stimuli (e.g., human vs. animal faces). It is still not clear, however, how the brain forms such classes, especially when introduced with new or changing environments. To isolate a perception-based mechanism underlying class representation, we studied unsupervised classification of an incoming stream of simple images. Classification patterns were clearly affected by stimulus frequency distribution, although subjects were unaware of this distribution. There was a common bias to locate class centers near the most frequent stimuli and their boundaries near the least frequent stimuli. Responses were also faster for more frequent stimuli. Using a minimal, biologically based neural-network model, we demonstrate that a simple, self-organizing representation mechanism based on overlapping tuning curves and slow Hebbian learning suffices to ensure classification. Combined behavioral and theoretical results predict large tuning overlap, implicating posterior infero-temporal cortex as a possible site of classification.

Adult↗

PubMed2001

A case study on the choice, interpretation and checking of multilevel models for longitudinal binary outcomes.

Recent advances in statistical software have led to the rapid diffusion of new methods for modelling longitudinal data. Multilevel (also known as hierarchical or random effects) models for binary outcomes have generally been based on a logistic-normal specification, by analogy with earlier work for normally distributed data. The appropriate application and interpretation of these models remains somewhat unclear, especially when compared with the computationally more straightforward semiparametric or 'marginal' modelling (GEE) approaches. In this paper we pose two interrelated questions. First, what limits should be placed on the interpretation of the coefficients and inferences derived from random-effect models involving binary outcomes? Second, what diagnostic checks are appropriate for evaluating whether such random-effect models provide adequate fits to the data? We address these questions by means of an extended case study using data on adolescent smoking from a large cohort study. Bayesian estimation methods are used to fit a discrete-mixture alternative to the standard logistic-normal model, and posterior predictive checking is used to assess model fit. Surprising parallels in the parameter estimates from the logistic-normal and mixture models are described and used to question the interpretability of the so-called 'subject-specific' regression coefficients from the standard multilevel approach. Posterior predictive checks suggest a serious lack of fit of both multilevel models. The results do not provide final answers to the two questions posed, but we expect that lessons learned from the case study will provide general guidance for further investigation of these important issues.

Journal Article↗

PubMed2006

More complete gene silencing by fewer siRNAs: transparent optimized design and biophysical signature.

Highly accurate knockdown functional analyses based on RNA interference (RNAi) require the possible most complete hydrolysis of the targeted mRNA while avoiding the degradation of untargeted genes (off-target effects). This in turn requires significant improvements to target selection for two reasons. First, the average silencing activity of randomly selected siRNAs is as low as 62%. Second, applying more than five different siRNAs may lead to saturation of the RNA-induced silencing complex (RISC) and to the degradation of untargeted genes. Therefore, selecting a small number of highly active siRNAs is critical for maximizing knockdown and minimizing off-target effects. To satisfy these needs, a publicly available and transparent machine learning tool is presented that ranks all possible siRNAs for each targeted gene. Support vector machines (SVMs) with polynomial kernels and constrained optimization models select and utilize the most predictive effective combinations from 572 sequence, thermodynamic, accessibility and self-hairpin features over 2200 published siRNAs. This tool reaches an accuracy of 92.3% in cross-validation experiments. We fully present the underlying biophysical signature that involves free energy, accessibility and dinucleotide characteristics. We show that while complete silencing is possible at certain structured target sites, accessibility information improves the prediction of the 90% active siRNA target sites. Fast siRNA activity predictions can be performed on our web server at http://optirna.unl.edu/.

Artificial Intelligence↗

PubMed2004

Automatic classification of heartbeats using ECG morphology and heartbeat interval features.

A method for the automatic processing of the electrocardiogram (ECG) for the classification of heartbeats is presented. The method allocates manually detected heartbeats to one of the five beat classes recommended by ANSI/AAMI EC57:1998 standard, i.e., normal beat, ventricular ectopic beat (VEB), supraventricular ectopic beat (SVEB), fusion of a normal and a VEB, or unknown beat type. Data was obtained from the 44 nonpacemaker recordings of the MIT-BIH arrhythmia database. The data was split into two datasets with each dataset containing approximately 50,000 beats from 22 recordings. The first dataset was used to select a classifier configuration from candidate configurations. Twelve configurations processing feature sets derived from two ECG leads were compared. Feature sets were based on ECG morphology, heartbeat intervals, and RR-intervals. All configurations adopted a statistical classifier model utilizing supervised learning. The second dataset was used to provide an independent performance assessment of the selected configuration. This assessment resulted in a sensitivity of 75.9%, a positive predictivity of 38.5%, and a false positive rate of 4.7% for the SVEB class. For the VEB class, the sensitivity was 77.7%, the positive predictivity was 81.9%, and the false positive rate was 1.2%. These results are an improvement on previously reported results for automated heartbeat classification systems.

Algorithms↗

PubMed1993

The object-line inferiority effect in pigeons.

Eight pigeons (Columba livia) were trained to discriminate between diagonal lines presented alone or embedded in a redundant L-shape right-angle form. The stimuli were white and were presented in an environment that was otherwise totally dark. Numerous experiments done with human observers have shown a strong superiority effect when the diagonal lines are embedded in redundant contexts. However, in Experiment 1 of the present study, the pigeons discriminated significantly better between the two diagonal lines when presented alone than when they were embedded in the right-angle context. In order to check on the possibility that these results were restricted to the semi-Ganzfeld condition of Experiment 1, a second experiment was done with black stimuli presented on a white background. Results of Experiment 2 also showed a strong distractor effect. The results of the present experiments confirmed the predictions of the Heinemann and Chase model of pattern recognition by pigeons.

Animals↗

PubMed1999

A measure of medical instructional quality in ambulatory settings: the MedIQ.

BACKGROUND AND OBJECTIVES: Concerns exist about the quality of medical student training outside the academic medical center. Yet, measures of this quality are lacking. This study introduces an instrument to measure instructional activities in primary care settings, based on a learner-centered model. The study also examines the instrument's ability to predict specific learner outcomes. METHODS: The MedIQ is a 25-item instrument designed to assess preceptor activities, environmental interactions, learning opportunities, and learner involvement in patient care. The MedIQ was administered in third-year generalist clerkships at one medical school, and the results were compared to extant measures of precepting effectiveness, student grades, specialty choice, and National Board of Medical Examiners scores. RESULTS: The results revealed strong reliability for all scales, moderate construct validity, and weak criterion-related validity. Additionally, some scores predicted specialty choice. CONCLUSION: The MedIQ is a promising measure of instructional quality in ambulatory medical settings.

Ambulatory Care↗

PubMed2003

A model for tool-use traditions in primates: implications for the coevolution of culture and cognition.

Inspired by the demonstration that tool-use variants among wild chimpanzees and orangutans qualify as traditions (or cultures), we developed a formal model to predict the incidence of these acquired specializations among wild primates and to examine the evolution of their underlying abilities. We assumed that the acquisition of the skill by an individual in a social unit is crucially controlled by three main factors, namely probability of innovation, probability of socially biased learning, and the prevailing social conditions (sociability, or number of potential experts at close proximity). The model reconfirms the restriction of customary tool use in wild primates to the most intelligent radiation, great apes; the greater incidence of tool use in more sociable populations of orangutans and chimpanzees; and tendencies toward tool manufacture among the most sociable monkeys. However, it also indicates that sociable gregariousness is far more likely to produce the maintenance of invented skills in a population than solitary life, where the mother is the only accessible expert. We therefore used the model to explore the evolution of the three key parameters. The most likely evolutionary scenario is that where complex skills contribute to fitness, sociability and/or the capacity for socially biased learning increase, whereas innovative abilities (i.e., intelligence) follow indirectly. We suggest that the evolution of high intelligence will often be a byproduct of selection on abilities for socially biased learning that are needed to acquire important skills, and hence that high intelligence should be most common in sociable rather than solitary organisms. Evidence for increased sociability during hominin evolution is consistent with this new hypothesis.

Animals↗

PubMed2000

Multivariate predictive relationship between kinematic and functional activation patterns in a PET study of visuomotor learning.

Imaging studies of visuomotor learning have reported practice-related activation in brain regions mediating sensorimotor functions. However, development and testing of functional motor learning models, based on the relationship between imaging and behavioral measures, is complicated by the multidimensional nature of motoric control. In the present study, multivariate techniques were used to analyze [15O]water PET and kinematic correlates of learning in a visuomotor tracing task. Fourteen subjects traced a geometric form over a series of eight tracing trials, preceded and followed by baseline trials in which they passively viewed the geometric form. Simultaneous evaluation of multiple behavioral measures indicated that performance improvement was most strongly associated with a global performance measure and least strongly associated with measures of fine motor control. Results of three independent analytic techniques (i.e., intertrial correlation matrices, power function modeling, iterative canonical variate analysis) indicated that imaging and behavioral measures were most closely related on early learning trials. Performance improvement was associated with covarying increases in normalized activity among superior parietal, postcentral gyrus, and premotor regions and covarying decreases in normalized activity among cerebellar, inferior parietal, pallidal, and medial occipital regions. These findings suggest that performance improvement may be associated with increased activation in neural systems previously implicated in visually guided reaching and decreased activation in neural systems previously implicated in attentive visuospatial processing.

Adult↗

PubMed2002

When language affects cognition and when it does not: an analysis of grammatical gender and classification.

The focus of this work was on the relation between grammatical gender and categorization. In one set of studies, monolingual English-, Spanish-, French-, and German-speaking children and adults assigned male and female voices to inanimate objects. Results from Spanish and French speakers indicated effects of grammatical gender on classification; results from German speakers did not. A connectionist model simulated the contradicting findings. The connectionist networks were also used to investigate which aspect of grammatical gender was responsible for the different pattern of findings. The predictions from the connectionist simulations were supported by the results from an artificial language-learning task. The results from this work demonstrate how connectionist networks can be used to identify the differences between languages that affect categorization.

Adult↗

PubMed1989

An on-line investigation of prototype and exemplar strategies in classification.

Although prototype- and exemplar-based models of categorization are very different in character, they have proved difficult to distinguish experimentally. The research described here presents a priming technique for assessing the type of information retrieved at the moment that a categorization decision is made. This technique avoids many of the problems inherent in the standard paradigms. Data from six experiments are presented that demonstrate the usefulness of the technique and also address basic questions about the categorization process. Results bolster previous suggestions that categorization strategies may be mixed within a single experimental task and highlight the need for more specific predictions about when each strategy will come into play.

Adult↗

PubMed2001

A longitudinal study of the etiology of separation anxiety.

A longitudinal examination of the relation between separation experiences and the development of separation anxiety at age 3, 11 and 18 years was conducted. Three associative pathways were assessed. Conditioning events were not related to separation anxiety at age 3. Vicarious learning (modelling) in middle childhood (age 9 years) was the conditioning variable most strongly related to separation anxiety at age 11, accounting for 1.8% of the variance in symptoms. Separation experiences (hospitalisations) before the age of 9 were inversely correlated with separation anxiety at age 18. That is, more overnight hospital stays in childhood were related to less separation anxiety in late adolescence. However, none of these conditioning correlates remained significant predictors of separation anxiety in adjusted regression models. In contrast, certain "planned" separations in early-mid childhood were associated with lower levels of separation anxiety at later ages. Generally, the findings were consistent with predictions from the non-associative theory of fear acquisition. That vicarious learning processes appeared to modulate, albeit to a minor degree, the expression of separation anxiety during mid-late childhood suggests that there may be critical periods during which some individuals are susceptible to the interactive effects of both associative and non-associative processes. These findings serve to illustrate the complexity of fear acquisition, the relevance of developmental factors and the likely interplay between associative and non-associative processes in the etiology of fear and anxiety.

Adolescent↗

PubMed1994

KADIS: model-aided education in type I diabetes. Karlsburg Diabetes Management System.

Education and training in self-management of blood glucose control has become a permanent task for all people involved in the care of diabetic patients. Since this may be facilitated by applying state-of-the-art information technology, we have developed the decision support system KADIS (Karlsburg Diabetes Management System). It comprises computer-aided tools for (1) the evaluation (selection, aggregation, storage, statistics, graphics) of therapeutic data, e.g. from patients' logbooks, and (2) the simulation of daily profiles of glycaemia and insulinaemia on the basis of a mathematical model of the glucose-insulin regulatory system, parameters of which can be adapted to the characteristics of individual patients. The latter tool allows the patient to predict his response to any modification in the therapeutic regime and to learn how variations in timing, formulation and doses of insulin, in carbohydrate equivalents and absorption characteristics of meals, and in exercise may influence the daily pattern in glycaemia. This procedure has been well accepted as an educational tool by those patients who were 'self-managing' their metabolic control.

Computer Graphics↗

PubMed2026

SurvGRN: a multi-feature fusion framework for bladder cancer survival prediction.

Bladder cancer survival outcomes exhibit significant heterogeneity, influenced by multifaceted factors. While digital pathology-based survival models leveraging artificial intelligence show promise, they often overlook complementary data sources. Conversely, imaging lacks cellular detail, and genomics/proteomics entail complexity and cost. To integrate multidimensional data for enhanced survival prediction, we propose SurvGRN, a multi-feature fusion framework. SurvGRN synergistically combines clinical variables, transcriptomics, and digital pathology slides using a gated residual network architecture. Pathological features are extracted via multiple instance learning, while clinical and transcriptomic data are processed as static inputs. These features are dynamically fused using a long short-term memory (LSTM) network for comprehensive survival risk assessment. Evaluated on 400 bladder cancer patients, SurvGRN significantly outperformed existing methods: improving the C-index by 12.6% over DeepMISL; 20.6% and 7.1% over graph-based models (DeepGraphConv and Patch-GCN); and 5.4% and 4.0% over attention-based approaches (Surformer and HVTSurv). Ablation studies confirmed the contributions of pathology features (extracted via ResNet-50 pre-trained on bladder tissue), clinical/transcriptomic data, and the LSTM fusion. SurvGRN also enabled significant stratification of patients into distinct risk cohorts. This work demonstrates that holistic integration of multi-source data through tailored fusion architectures substantially improves bladder cancer survival prediction.

bladder cancer↗