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 1,351 records · Page 75Linked to original sources

PubMed1988

Value of static pulmonary compliance in predicting mortality in patients with acute respiratory failure.

The measurement of static pulmonary compliance by means of pressure-volume curves is an useful tool in the management of patients with acute respiratory failure. On admission to our Intensive Care Unit, we calculated the static pulmonary compliance values from the pressure-volume curve in 55 severe acute respiratory failure patients under mechanical ventilation. A predictive model, based on static pulmonary compliance measured on the deflation of the pressure-volume curve and patients' age, was developed using a function derived by stepwise discriminant analysis. Thirty-six of the 40 patients (90%) in the learning sample and 12 of the 15 patients (80%) in the test sample were correctly classified into survivors or non-survivors. The overall classification demonstrated an accuracy of 88% (48 of 55 patients). We conclude that the value of static pulmonary compliance can be an useful prognostic factor in patients with severe acute respiratory failure.

Adolescent↗

PubMed2003

Language learning and innateness: some implications of Compounds Research.

In noun compounds in English, the modifying noun may be singular (mouse-eater) or an irregularly inflected plural (mice-eater), but regularly inflected plurals are dispreferred (*rats-eater). This phenomenon has been taken as strong evidence for dual-mechanism theories of lexical representations, which hold that regular (rule-governed) and irregular (exception) items are generated by qualitatively different and innately specified mechanisms. Using corpus analyses, behavioral studies, and computational modeling, we show that the rule-versus-exceptions approach makes a number of incorrect predictions. We propose a new account in which the acceptability of modifiers is determined by a constraint satisfaction process modulated by semantic, phonological, and other factors. The constraints are acquired by the child via general purpose learning algorithms, based on noun compounds and other constructions in the input. The account obviates the regular/irregular dichotomy while simultaneously providing a superior account of the data.

Child↗

PubMed1998

Induction of relational schemas: common processes in reasoning and complex learning.

Five experiments were performed to test whether participants induced a coherent representation of the structure of a task, called a relational schema, from specific instances. Properties of a relational schema include: An explicit symbol for a relation, a binding that preserves the truth of a relation, potential for higher-order relations, omnidirectional access, potential for transfer between isomorphs, and ability to predict unseen items in isomorphic problems. However relational schemas are not necessarily coded in abstract form. Predictions from relational schema theory were contrasted with predictions from configural learning and other nonstructural theories in five experiments in which participants were taught a structure comprised of a set of initial-state,operator-->end-state instances. The initial-state,operator pairs were presented and participants had to predict the correct end-state. Induction of a relational schema was achieved efficiently by adult participants as indicated by ability to predict items of a new isomorphic problem. The relational schemas induced showed the omnidirectional access property, there was efficient transfer to isomorphs, and structural coherence had a powerful effect on learning. The "learning to learn" effect traditionally associated with the learning set literature was observed, and the long-standing enigma of learning set acquisition is explained by a model composed of relational schema induction and structure mapping. Performance was better after reversal of operators than after shift to an alternate structure, even though the latter entailed more overlap with previously learned tasks in terms of the number of configural associations that were preserved. An explanation for the reversal shift phenomenon in terms of induction and mapping of a relational schema is proposed. The five experiments provided evidence supporting predictions from relational schema theory, and no evidence was found for configural or nonstructural learning theories.

Adult↗

PubMed2006

Prediction of the concentration of chlorophyll-a for Liuhai urban lakes in Beijing City.

The weekly water quality monitor data of Liuhai lakes between April 2003 and November 2004 in Beijing City were used as an example to build an artificial neural networks (ANN) model and a multi-varieties regression model respectively for predicting the fresh water algae bloom. The different predicted abilities of the two methods in Liuhai lakes were compared. A principle analysis method was first used to select the input variables of the models to avoid the phenomenon of collinearity in the data. The results showed that the input variables for the artificial neural networks were T, TP, transparency(SD), DO, chlorophyll-a (Chl-a), pH and the output variable was Chl-a. A three layer Levenberg-Marguardt feed forward learning algorithm in ANN was used to model the eutrophication process of Liuhai lakes. 20 nodes in hidden layer and 1 node of output for the ANN model had been optimized by trial and error method. A sensitivity analysis of the input variables was performed to evaluate their relative significance in determining the predicted values. The correlation coefficient between predicted value and observed value in all data and in test data were 0.717 and 0.816 respectively in the artificial neural networks. The stepwise regression method was used to simulate the linear relation between Chl-a and temperature, of which the correlation coefficient was 0.213. By comparing the results of the two models, it was found that neural network models were able to simulate non-linear behavior in the water eutrophication process of Liuhai lakes reasonably and could successfully estimate some extreme values from calibration and test data sets.

China↗

PubMed2004

Prediction of left ventricular ejection fraction in patients with coronary artery disease based on an analysis of perfusion patterns at rest. Assessment by an artificial neural network.

BACKGROUND: In CAD, left ventricular function depends on the condition of myocardial perfusion, hence it may be presumed that blood flow abnormalities may enable the LVEF to be predicted. The aim of the study was to apply an Artificial Neural Network (ANN) to investigate the relationships between myocardial perfusion and LVEF, measured simultaneously. MATERIAL AND METHODS: gSPECT examinations were performed in 95 patients with CAD, divided into training (n = 50) and testing (n = 45) groups. using the acquired data, in each subject the LVEF was calculated and a perfusion polar map was constructed and divided into 25 segments. Based on results obtained in the training group, a characteristic configuration of segments was defined, with features enabling differentiation between the individual subjects of that group. The set of those segments, as well as the corresponding LVEF values enabled the optimum network architecture to be constructed and trained. The trained ANN was verified by application to the testing group. RESULTS: Using the above-described procedure, 15 polar map segments were defined which enabled the patients of the training group to be differentiated sufficiently enough to make their further recognition possible. The optimal network structure consisting 25 neurons was obtained by comparing the activity in those segments in individual subjects with corresponding LVEF values. Based on the above model, the obtained network was able to reproduce learning data (r = 0.832; learning error = 4.84%) and to apply the gained knowledge to the testing cases (r = 0.786; testing error = 4.99%). CONCLUSIONS: The obtained network can generalise learned information. To predict LVEF, some polar map segments should be excluded from the analysis. Erroneous LVEF prediction may occur resulting mainly from conditions independent from perfusion abnormalities.

Artificial Intelligence↗

PubMed2000

Individual differences in pavlovian autoshaping of lever pressing in rats predict stress-induced corticosterone release and mesolimbic levels of monoamines.

Pavlovian autoshaping CRs are directed and reflexive consummatory responses targeted at objects repeatedly paired with rewarding substances. To evaluate the hypothesis that autoshaping may provide an animal learning model of vulnerability to drug abuse, this study relates individual differences in lever-press autoshaping CR performance in rats to stress-induced corticosterone release and tissue monoamine levels in the mesolimbic dopamine tract. Long-Evans rats (n = 14) were given 20 sessions of Pavlovian autoshaping training wherein the insertion of a retractable lever CS was followed by the response-independent presentation of food US. Large between-subjects differences in lever-press autoshaping CR performance were observed, with group high CR frequency (n = 5) performing many more lever press CRs than group low CR frequency (n = 9). Tail-blood samples were obtained before and after the 20th autoshaping session, then 24 h later the rats were sacrificed and dissection yielded tissue samples of nucleus accumbens (NAC), prefrontal cortex (PFC), caudate putamen (CP), and ventral tegmental area (VTA). Serum levels of postsession corticosterone were elevated in group high CR frequency. HPLC revealed that group high CR frequency had higher tissue levels of dopamine and DOPAC in NAC, lower levels of DOPAC/DA turnover in CP, and lower levels of 5-HIAA and lower 5-HIAA/5-HT turnover in VTA. The neurochemical profile of rats that perform more autoshaping CRs share some features of vulnerability to drug abuse.

Animals↗

PubMed2001

Excitotoxic lesions of the hippocampus leave sensory preconditioning intact: implications for models of hippocampal function.

Learning about contextual cues is markedly disrupted in rats with hippocampal lesions. One analysis of this disruption is that it reflects a general failure to form associations between the elements of complex events. A straightforward prediction of this analysis is that sensory preconditioning will be disrupted by hippocampal lesions. This prediction was assessed by presenting rats with flavored solutions composed of 2 elements (A and X) before X was paired with an injection of the emetic, lithium chloride. A subsequent test revealed that rats were less willing to consume Solution A than they were to consume a control solution, B. This was true of rats with sham lesions and those with excitotoxic lesions of hippocampus. These findings fail to support the proposition that the hippocampus-dependent deficit in contextual conditioning is due to a general disruption to the process of associating the elements of complex events.

Animals↗

PubMed2006

Semi-supervised analysis of gene expression profiles for lineage-specific development in the Caenorhabditis elegans embryo.

MOTIVATION: Gene expression profiling is a powerful approach to identify genes that may be involved in a specific biological process on a global scale. For example, gene expression profiling of mutant animals that lack or contain an excess of certain cell types is a common way to identify genes that are important for the development and maintenance of given cell types. However, it is difficult for traditional computational methods, including unsupervised and supervised learning methods, to detect relevant genes from a large collection of expression profiles with high sensitivity and specificity. Unsupervised methods group similar gene expressions together while ignoring important prior biological knowledge. Supervised methods utilize training data from prior biological knowledge to classify gene expression. However, for many biological problems, little prior knowledge is available, which limits the prediction performance of most supervised methods. RESULTS: We present a Bayesian semi-supervised learning method, called BGEN, that improves upon supervised and unsupervised methods by both capturing relevant expression profiles and using prior biological knowledge from literature and experimental validation. Unlike currently available semi-supervised learning methods, this new method trains a kernel classifier based on labeled and unlabeled gene expression examples. The semi-supervised trained classifier can then be used to efficiently classify the remaining genes in the dataset. Moreover, we model the confidence of microarray probes and probabilistically combine multiple probe predictions into gene predictions. We apply BGEN to identify genes involved in the development of a specific cell lineage in the C. elegans embryo, and to further identify the tissues in which these genes are enriched. Compared to K-means clustering and SVM classification, BGEN achieves higher sensitivity and specificity. We confirm certain predictions by biological experiments. AVAILABILITY: The results are available at http://www.csail.mit.edu/~alanqi/projects/BGEN.html.

Algorithms↗

PubMedDate not supplied

Predicting student attrition in a baccalaureate curriculum.

A longitudinal study of enrollment data of the Universtiy of Wisconsin--Madison School of Nursing identified several trends in student attrition over a six-year period and formed the basis for the development of a model to predict student attrition. The model proposed to satisfy the need for more precise measures of probable student retention and attrition for counseling and planning purposes in the school's new curriculum. To arrive at a means for predicting student attrition, scores of random samples of continuing and dropout student on achievement, learning style, and psychologial variables obtained during the first three years of the new curriculum were examined using a discriminant analysis technique. A significant differentiation between continuing and dropout students (p less than .000) resulted and allowed a predictive function to be developed. Using a conservative method of identification with student scores transformed by this function, predictions of successful and nonsuccessful students were obtained.

Curriculum↗

PubMed2002

Comparing prototype-based and exemplar-based accounts of category learning and attentional allocation.

Exemplar theory was motivated by research that often used D. L. Medin and M. M. Schaffer's (1978) 5/4 stimulus set. The exemplar model has seemed to fit categorization data from this stimulus set better than a prototype model can. Moreover, the exemplar model alone predicts a qualitative aspect of performance that participants sometimes show. In 2 experiments, the authors reexamined these findings. In both experiments, a prototype model fit participants' performance profiles better than an exemplar model did when comparable prototype and exemplar models were used. Moreover, even when participants showed the qualitative aspect of performance, the exemplar model explained it by making implausible assumptions about human attention and effort in categorization tasks. An independent assay of participants' attentional strategies suggested that the description the exemplar model offers in such cases is incorrect. A review of 30 uses of the 5/4 stimulus set in the literature reinforces this suggestion.

Attention↗

PubMed1993

Negative priming in associative learning: evidence from a serial-conditioning procedure.

Three experiments investigated the suggestion that a predicted or primed stimulus commands less processing and consequently elicits a weaker conditioned response (CR) than a stimulus that is not primed. In each experiment rats received initial training in which the presentation of each of 2 serial compounds, A-X and B-Y, was followed by the delivery of food. Subsequently, X's capacity to elicit the CR, approaching the site of food delivery, was assessed when X was preceded by Stimulus A (i.e., primed) or was presented after Stimulus B. Stimulus X elicited a more vigorous response when it was presented after B than when it followed A. These results show that the ability of one event to elicit its CR is reduced if its presentation has been predicted by some other event. This negative priming effect supports one aspect of Wagner's (1981) model of Pavlovian conditioning.

Animals↗

PubMed2026

Fairness-aware supervised hierarchical contrastive semantic learning for sexual dimorphism analysis.

MOTIVATION: Sexual dimorphism is a fundamental biological determinant driving systematic differences in disease susceptibility, progression, and clinical outcomes. However, current sex-combined AI-based genomic models often exhibit algorithmic bias and fail to capture these sex-specific mechanisms, creating a critical barrier to unbiased precision medicine. Ensuring fairness in the context of sexual dimorphism requires understanding and addressing the distinct biological mechanisms functioning in each sex, rather than focusing solely on equalizing predictive performance. RESULTS: We propose a fairness-aware supervised hierarchical contrastive learning approach, called FairHICON, to discover unbiased sex-common and sex-specific predictive features. Evaluations on cancer and asthma transcriptomic datasets demonstrate that FairHICON significantly outperforms state-of-the-art benchmarks, improving predictive performance by up to 9% while effectively reducing the performance gap between male and female sexes. Furthermore, prognostic validation confirms that the identified sex-specific pathways stratify patient survival significantly better within their corresponding sex groups. This validates FairHICON to elucidate the molecular heterogeneity of sexual dimorphism, advancing inclusive precision medicine. AVAILABILITY AND IMPLEMENTATION: The source code and data is available at https://github.com/datax-lab/FairHICON.

Sex Characteristics↗

PubMed1991

Conditioned reinforcement value and choice.

The delay-reduction hypothesis of conditioned reinforcement states that the reinforcing value of a food-associated stimulus is determined by the delay to primary reinforcement signaled by the onset of the stimulus relative to the average delay to primary reinforcement in the conditioning situation. In contrast, most contemporary models of conditioned reinforcement strength posit that the reinforcing strength of a stimulus is some simple function only of the delay to primary reinforcement in the presence of stimulus. The delay-reduction hypothesis diverges from other conditioned reinforcement models in that it predicts that a fixed-duration food-paired stimulus will have different reinforcing values depending on the frequency of its presentation. In Experiment 1, pigeons' key pecks were reinforced according to concurrent-chains schedules with variable-interval 10-second and variable-interval 20-second terminal-link schedules. The initial-link schedule preceding the shorter terminal link was always variable-interval 60 seconds, and the initial-link schedule requirement preceding the longer terminal link was varied between 1 second and 60 seconds across conditions. In Experiment 2, the initial-link schedule preceding the longer of two terminal links was varied for each of three groups of pigeons. The terminal links of the concurrent chains for the three groups were variable-interval 10 seconds and 20 seconds, variable-interval 10 seconds and 30 seconds, and variable-interval 30 seconds and 50 seconds. In both experiments, preference for the shorter terminal link was either a bitonic function or an inverse function of the initial-link schedule preceding the longer terminal-link schedule. Consistent with the predictions of the delay-reduction hypothesis, the relative values of the terminal-link stimuli changed as a function of the overall frequency of primary reinforcement. Vaughan's (1985) melioration model, which was shown to be formally similar to Squires and Fantino's (1971) delay-reduction model, can be modified so as to predict these results without changing its underlying assumptions.

Animals↗

PubMed2005

The centralization phenomenon and fear-avoidance beliefs as prognostic factors for acute low back pain: a preliminary investigation involving patients classified for specific exercise.

STUDY DESIGN: Secondary analysis of a prospective cohort of patients with acute low back pain (LBP). OBJECTIVES: To determine if the centralization phenomenon and fear-avoidance beliefs predict measurement of pain and disability 6 months after entering the study. BACKGROUND: The centralization phenomenon and fear-avoidance are predictive of future pain and disability. However, previous prognostic studies have not routinely included both measures in homogenous subgroups of patients with acute LBP. METHODS AND MEASURES: Patients completed self-report questionnaires and were evaluated and treated with treatment-based classification guidelines. Only the patients classified for specific exercise were included in this analysis (n = 28). Measures of disability and pain intensity were reassessed at 6 months by mail. Separate hierarchical regression models predicted measures of disability and pain intensity with the centralization phenomenon, fear-avoidance beliefs, and prespecified covariates. RESULTS: There were no significant differences in duration of symptoms, fear-avoidance beliefs, and history of LBP based on the centralization phenomenon (P > .05). Patients reporting the centralization phenomenon were significantly more likely to have leg pain (P < .01). A regression model including initial disability, the centralization phenomenon, and fear-avoidance beliefs about work significantly predicted 6-month disability, explaining 49% of the total variance (P < .001). A regression model that included initial pain intensity and the centralization phenomenon significantly predicted 6-month pain intensity, explaining 29% of the total variance (P < .016). These factors also appeared to be clinically meaningful predictors of outcome, but lacked precision for immediate use in clinical settings. The following covariates were not included in the final regression models: presence of leg pain, history of LBP, and duration of LBP. CONCLUSIONS: Baseline elevation in fear-avoidance beliefs about work and lack of centralization phenomenon predicted higher disability. Baseline lack of centralization phenomenon predicted higher pain intensity. These results can only be generalized to patients with acute LBP classified for specific exercise. It will be necessary to independently validate these prediction models before they can be implemented in clinical settings.

Acute Disease↗

PubMed2002

Fruits and vegetables categorized: an application of the generalized context model.

In the study reported in this paper, we investigated the categorization of well-known and novel food items in the categories fruits and vegetables. Predictions based on Nosofsky's (1984,1986) generalized context model (GCM), on a multiplicative-similarity prototype model, and on an instantiation model as applied in Storms, De Boeck, and Ruts (2001) were compared. Despite suggestions in the literature that prototype models predict categorization from large categories better than exemplar models do, our results showed that the exemplar-based GCM yielded clearly better predictions than did a (multiplicative-similarity) prototype model.

Adult↗

PubMed1990

The perceptual buildup of three-dimensional structure from motion.

We present a set of psychophysical experiments that measure the accuracy of perceived three-dimensional (3-D) structure derived from relative motion in the changing two-dimensional image. The experiments are motivated in part by a computational model proposed by Ullman (1984), called the incremental rigidity scheme, in which an accurate 3-D structure is built up incrementally, by considering images of moving objects over an extended time period. Our main conclusions are: First, the human visual system can derive an accurate model of the relative depths of moving points, even in the presence of noise in their image positions; second, the accuracy of the 3-D model improves with time, eventually reaching a plateau; and third, the 3-D structure currently perceived appears to depend on previous 3-D models. Through computer simulations, we relate the results of our psychophysical experiments with the predictions of Ullman's model.

Adult↗

PubMed2003

The case for an internal dynamics model versus equilibrium point control in human movement.

The equilibrium point hypothesis (EPH) was conceived as a means whereby the central nervous system could control limb movements by a relatively simple shift in equilibrium position without the need to explicitly compensate for task dynamics. Many recent studies have questioned this view with results that suggest the formation of an internal dynamics model of the specific task. However, supporters of the EPH have argued that these results are not incompatible with the EPH and that there is no reason to abandon it. In this study, we have tested one of the fundamental predictions of the EPH, namely, equifinality. Subjects learned to perform goal-directed wrist flexion movements while a motor provided assistance in proportion to the instantaneous velocity. It was found that the subjects stopped short of the target on the trials where the magnitude of the assistance was randomly decreased, compared to the preceding control trials (P = 0.003), i.e. equifinality was not achieved. This is contrary to the EPH, which predicts that final position should not be affected by external loads that depend purely on velocity. However, such effects are entirely consistent with predictions based on the formation of an internal dynamics model.

Adult↗

PubMed2006

Remembering source evidence from associatively related items: explanations from a global matching model.

The authors tested source memory across three conditions, one in which 3 strongly associated primes of a target word were presented in the same source as the target, one in which primes were presented in a different source than the target, and one in which no associates of targets were encoded. In the first 2 experiments, target source memory increased in the same-prime condition and decreased in the different-prime condition relative to the no-prime condition. In Experiment 3, the different-prime condition created the illusion that target words had been presented in both sources at encoding. The MINERVA 2 model (D. L. Hintzman, 1988) was able to predict these effects by basing source decisions on the global match of source-specific retrieval probes to all of the items in the memory set.

Adolescent↗