Computational Learning Model Research Articles

Human motor learning dynamics in high-dimensional tasks.

Conventional approaches to enhance movement coordination, such as providing instructions and visual feedback, are often inadequate in complex motor tasks with multiple degrees of freedom (DoFs). To effectively address coordination deficits in such complex motor systems, it becomes imperative to develop interventions grounded in a model of human motor learning; however, modeling such learning processes is challenging due to the large DoFs. In this paper, we present a computational motor learning model that leverages the concept of motor synergies to extract low-dimensional learning representations in the high-dimensional motor space and the internal model theory of motor control to capture both fast and slow motor learning processes. We establish the model's convergence properties and validate it using data from a target capture game played by human participants. We study the influence of model parameters on several motor learning trade-offs such as speed-accuracy, exploration-exploitation, satisficing, and flexibility-performance, and show that the human motor learning system tunes these parameters to optimize learning and various output performance metrics.

Lack of optimistic bias during social evaluation learning reflects reduced positive self-beliefs in depression and social anxiety, but via distinct mechanisms

Processing social feedback optimistically may maintain positive self-beliefs and stable social relationships. Conversely, a lack of this optimistic bias in depression and social anxiety may perpetuate negative self-beliefs and maintain symptoms. Research investigating this mechanism is scarce, however, and the mechanisms by which depressed and socially anxious individuals respond to social evaluation may also differ. Using a range of computational approaches in two large datasets (mega-analysis of previous studies, n = 450; pre-registered replication study, n = 807), we investigated how depression (PHQ-9) and social anxiety (BFNE) symptoms related to social evaluation learning in a computerized task. Optimistic bias (better learning of positive relative to negative evaluations) was found to be negatively associated with depression and social anxiety. Structural equation models suggested this reflected a heightened sensitivity to negative social feedback in social anxiety, whereas in depression it co-existed with a blunted response to positive social feedback. Computational belief-based learning models further suggested that reduced optimism was driven by less positive trait-like self-beliefs in both depression and social anxiety, with some evidence for a general blunting in belief updating in depression. Recognizing such transdiagnostic similarities and differences in social evaluation learning across disorders may inform approaches to personalizing treatment.

Transfer learning may explain pigeons’ ability to detect cancer in histopathology

Pigeons' unexpected competence in learning to categorize unseen histopathological images has remained an unexplained discovery for almost a decade (Levensonet al2015PLoS One10e0141357). Could it be that knowledge transferred from their bird's-eye views of the earth's surface gleaned during flight contributes to this ability? Employing a simulation-based verification strategy, we recapitulate this biological phenomenon with a machine-learning analog. We model pigeons' visual experience during flight with the self-supervised pre-training of a deep neural network on BirdsEyeViewNet; our large-scale aerial imagery dataset. As an analog of the differential food reinforcement performed in Levensonet al's study 2015PLoS One10e0141357), we apply transfer learning from this pre-trained model to the same Hematoxylin and Eosin (H&E) histopathology and radiology images and tasks that the pigeons were trained and tested on. The study demonstrates that pre-training neural networks with bird's-eye view data results in close agreement with pigeons' performance. These results support transfer learning as a reasonable computational model of pigeon representation learning. This is further validated with six large-scale downstream classification tasks using H&E stained whole slide image datasets representing diverse cancer types.

Adaptive Evolutionary Computing Ensemble Learning Model for Sentiment Analysis

Standard machine learning and deep learning architectures have been widely used in the field of sentiment analysis, but their performance is unsatisfactory if the input texts are short (e.g., social media posts). Specifically, the accuracy of standard machine learning methods crucially depends on the richness and completeness of the features used to represent the texts, and in the case of short messages, it is often difficult to obtain high-quality features. Conversely, methods based on deep learning can achieve better expressiveness, but these methods are computationally demanding and often suffer from over-fitting. This paper proposes a new adaptive evolutionary computational integrated learning model (AdaECELM) to overcome the problems encountered by traditional machine learning and deep learning models in sentiment analysis for short texts. AdaECELM consists of three phases: feature selection, sub classifier training, and global integration learning. First, a grid search is used for feature extraction and selection of term frequency-inverse document frequency (TF-IDF). Second, cuckoo search (CS) is introduced to optimize the combined hyperparameters in the sub-classifier support vector machine (SVM). Finally, the training set is divided into different feature subsets for sub-classifier training, and then the trained sub-classifiers are integrated and learned using the AdaBoost integrated soft voting method. Extensive experiments were conducted on six real polar sentiment analysis data sets. The results show that the AdaECELM model outperforms the traditional ML comparison methods according to evaluation metrics such as accuracy, precision, recall, and F1-score in all cases, and we report an improvement in accuracy exceeding 4.5%, the second-best competitor.

Tailoring therapies: a frontier approach to pancreatic cancer with AI-driven multiomics profiling

<p>Pancreatic cancer is often diagnosed at an advanced stage when treatment options are limited. Being one of the deadliest cancers that mandates longer medication and treatment phases, there is an inevitable need to have the knowledge of drug response of anti-pancreatic cancer drugs before it is recommended for a patient. AI-driven drug response prediction has proven potential to personalize treatment strategies, improve therapeutic outcomes, and reduce adverse effects and treatment costs for cancer patients. In this research work, we have accounted for the use of different drug descriptors and their core structures known as scaffolds along with three cell line features, chromatin profiling, reverse phase protein array, and metabolomics data to build a feature engineered dataset for drug response prediction tested on various computational learning models. The 53 unique drugs against 18 unique pancreatic cancer cell lines were taken as the raw dataset. The initial dataset having a large dimension was feature selected using an ensemble method derived from five different techniques. The dataset was evaluated on various computational methods and an accuracy of 89% was achieved using the TabNet architecture. Furthermore, the common scaffolds that were persistently found among the drugs that possess high IC50-valued drug clusters were also recorded.</p>

Reinforcement learning of altruistic punishment differs between cultures and across the lifespan.

Altruistic punishment is key to establishing cooperation and maintaining social order, yet its developmental trends across cultures remain unclear. Using computational reinforcement learning models, we provided the first evidence of how social feedback dynamically influences group-biased altruistic punishment across cultures and the lifespan. Study 1 (n = 371) found that Chinese participants exhibited higher learning rates than Americans when socially incentivized to punish unfair allocations. Additionally, Chinese adults showed slower learning and less exploration when punishing ingroups than outgroups, a pattern absent in American counterparts, potentially reflecting a tendency towards ingroup favoritism that may contribute to reinforcing collectivist values. Study 2 (n = 430, aged 12-52) further showed that such ingroup favoritism develops with age. Chinese participants' learning rates for ingroup punishment decreased from adolescence into adulthood, while outgroup rates stayed constant, implying a process of cultural learning. Our findings highlight cultural and age-related variations in altruistic punishment learning, with implications for social reinforcement learning and culturally sensitive educational practices promoting fairness and altruism.

Individual differences in autism-like traits are associated with reduced goal emulation in a computational model of observational learning

Individual differences in autism-like traits are associated with reduced goal emulation in a computational model of observational learning

Electrical brain activations in preadolescents during a probabilistic reward-learning task reflect cognitive processes and behavioral strategy.

Both adults and children learn through feedback which environmental events and choices are associated with higher probability of reward, an ability thought to be supported by the development of fronto-striatal reward circuits. Recent developmental studies have applied computational models of reward learning to investigate such learning in children. However, tasks and measures effective for assaying the cascade of reward-learning neural processes in children have been limited. Using a child-version of a probabilistic reward-learning task while recording event-related-potential (ERP) measures of electrical brain activity, this study examined key processes of reward learning in preadolescents (8-12 years old; n=30), namely: (1) reward-feedback sensitivity, as measured by the early-latency, reward-related, frontal ERP positivity, (2) rapid attentional shifting of processing toward favored visual stimuli, as measured by the N2pc component, and (3) longer-latency attention-related responses to reward feedback as a function of behavioral strategies (i.e., Win-Stay-Lose-Shift), as measured by the central-parietal P300. Consistent with our prior work in adults, the behavioral findings indicate preadolescents can learn stimulus-reward outcome associations, but at varying levels of performance. Neurally, poor preadolescent learners (those with slower learning rates) showed greater reward-related positivity amplitudes relative to good learners, suggesting greater reward-feedback sensitivity. We also found attention shifting towards to-be-chosen stimuli, as evidenced by the N2pc, but not to more highly rewarded stimuli as we have observed in adults. Lastly, we found the behavioral learning strategy (i.e., Win-Stay-Lose-Shift) reflected by the feedback-elicited parietal P300. These findings provide novel insights into the key neural processes underlying reinforcement learning in preadolescents.

Social learning in models and minds

After more than a century in which social learning was blackboxed by evolutionary biologists, psychologists and economists, there is now a thriving industry in cognitive neuroscience producing computational models of learning from and about other agents. This is a hugely positive development. The tools of computational cognitive neuroscience are rigorous and precise. They have the potential to prise open the black box. However, we argue that, from the perspective of a scientific realist, these tools are not yet being applied in an optimal way. To fulfil their potential, the shiny new methods of cognitive neuroscience need to be better coordinated with old-fashioned, contrastive experimental designs. Inferences from model complexity to cognitive complexity, of the kind made by those who favour lean interpretations of behaviour (‘associationists’), require social learning to be tested in challenging task environments. Inferences from cognitive complexity to social specificity, made by those who favour rich interpretations (‘mentalists’), call for non-social control experiments. A parsimonious model that fits current data is a good start, but carefully designed experiments are needed to distinguish models that tell us how social learning could be done from those that tell us how it is really done.

Neurocomputational Mechanisms Underlying Differential Reinforcement Learning From Wins and Losses in Obesity With and Without Binge Eating

BackgroundBinge-eating disorder (BED) is thought of as a disorder of cognitive control, but evidence regarding its neurocognitive mechanisms is inconclusive. Key limitations of previous research include a lack of consistent separation between effects of BED and obesity and a disregard for self-report evidence suggesting that neurocognitive alterations may emerge primarily in loss- or harm-avoidance contexts. MethodsTo address these gaps, in this longitudinal study we investigated behavioral flexibility and its underlying neurocomputational processes in reward-seeking and loss-avoidance contexts. Obese participants with BED, obese participants without BED, and healthy normal-weight participants (n = 96) performed a probabilistic reversal learning task during functional imaging, with different blocks focused on obtaining wins or avoiding losses. They were reinvited for a 6-month follow-up assessment. ResultsAnalyses informed by computational models of reinforcement learning showed that unlike obese participants with BED, obese participants without BED performed worse in the win than in the loss condition. Computationally, this was explained by differential learning sensitivities in the win versus loss conditions in the groups. In the brain, this was echoed in differential neural learning signals in the ventromedial prefrontal cortex per condition. The differences were subtle but scaled with BED symptoms, such that more severe BED symptoms were associated with increasing bias toward improved learning from wins versus losses. Across conditions, obese participants with BED switched more between choice options than healthy normal-weight participants. This was reflected in diminished representation of choice certainty in the ventromedial prefrontal cortex. ConclusionsOur study highlights the importance of distinguishing between obesity with and without BED to identify unique neurocomputational alterations underlying different styles of maladaptive eating behavior.

Mediated learning: A computational rendering of ketamine-induced symptoms.

This article explores the contribution of the double error dynamic asymptote computational associative learning model to understanding the role of mediated learning mechanisms in the generation of spurious associations, as those postulated to characterize schizophrenia. Three sets of simulations for mediated conditioning, mediated extinction, and a mediated enhancement of latent inhibition, a unique model prediction, are presented. For each set of simulations, a parameter that modulates the impact of associative memory retrieval and the dissipation of nonperceptual activated representations through the network was manipulated. The effect of this operation is analyzed and compared to ketamine-induced effects on associative memories and mediated learning. The model's potential to predict these effects and present a plausible error-correction associative mechanism is discussed in the context of animal models of schizophrenia. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Impairment in acquisition of conditioned fear in people with depressive symptoms.

Depression is one of the primary global public health issues, and there has been a dramatic increase in depression levels among young people over the past decade. The neuroplasticity theory of depression postulates that a malfunction in neural plasticity, which is responsible for learning, memory, and adaptive behavior, is the primary source of the disorder's clinical manifestations. Nevertheless, the impact of depression symptoms on associative learning remains underexplored. We used the differential fear conditioning paradigm to investigate the effects of depressive symptoms on fear acquisition and extinction learning. Skin conductance response (SCR) is an objective evaluation indicator, and ratings of nervousness, likeability, and unconditioned stimuli (US) expectancy are subjective evaluation indicators. In addition, we used associability generated by a computational reinforcement learning model to characterize the skin conductance response. The findings indicate that individuals with depressive symptoms exhibited significant impairment in fear acquisition learning compared to those without depressive symptoms based on the results of the skin conductance response. Moreover, in the discrimination fear learning task, the skin conductance response was positively correlated with associability, as estimated by the hybrid model in the group without depressive symptoms. Additionally, the likeability rating scores improved post-extinction learning in the group without depressive symptoms, and no such increase was observed in the group with depressive symptoms. The study highlights that individuals with pronounced depressive symptoms exhibit impaired fear acquisition and extinction learning, suggesting a possible deficit in associative learning. Employing the hybrid model to analyze the learning process offers a deeper insight into the associative learning processes of humans, thus allowing for improved comprehension and treatment of these mental health problems.

Nano Empowered Neural Network Based Adaptive Hybridized Computational Model for Simulated Word Learning and its Implications for Human Psychology

Nano Empowered Neural Network Based Adaptive Hybridized Computational Model for Simulated Word Learning and its Implications for Human Psychology

A-Cation-Dependent Excited State Charge Carrier Dynamics in Vacancy-Ordered Halide Perovskites: Insights from Computational and Machine Learning Models

A-Cation-Dependent Excited State Charge Carrier Dynamics in Vacancy-Ordered Halide Perovskites: Insights from Computational and Machine Learning Models

Training self-learning circuits for power-efficient solutions

As the size and ubiquity of artificial intelligence and computational machine learning models grow, the energy required to train and use them is rapidly becoming economically and environmentally unsustainable. Recent laboratory prototypes of self-learning electronic circuits, such as “physical learning machines,” open the door to analog hardware that directly employs physics to learn desired functions from examples at a low energy cost. In this work, we show that this hardware platform allows for an even further reduction in energy consumption by using good initial conditions and a new learning algorithm. Using analytical calculations, simulations, and experiments, we show that a trade-off emerges when learning dynamics attempt to minimize both the error and the power consumption of the solution—greater power reductions can be achieved at the cost of decreasing solution accuracy. Finally, we demonstrate a practical procedure to weigh the relative importance of error and power minimization, improving the power efficiency given a specific tolerance to error.

The social transmission of empathy relies on observational reinforcement learning

Theories of moral development propose that empathy is transmitted across individuals. However, the mechanisms through which empathy is socially transmitted remain unclear. Here, we combine computational learning models and functional MRI to investigate whether, and if so, how empathic and non-empathic responses observed in others affect the empathy of female observers. The results of three independent studies showed that watching empathic or non-empathic responses generates a learning signal that respectively increases or decreases empathy ratings of the observer. A fourth study revealed that the learning-related transmission of empathy is stronger when observing human rather than computer demonstrators. Finally, we show that the social transmission of empathy alters empathy-related responses in the anterior insula, i.e., the same region that correlated with empathy baseline ratings, as well as its functional connectivity with the temporoparietal junction. Together, our findings provide a computational and neural mechanism for the social transmission of empathy that accounts for changes in individual empathic responses in empathic and non-empathic social environments.

A Machine Learning Predictive Model of Bloodstream Infection in Hospitalized Patients.

The aim of the study was to build a machine learning-based predictive model to discriminate between hospitalized patients at low risk and high risk of bloodstream infection (BSI). A Data Mart including all patients hospitalized between January 2016 and December 2019 with suspected BSI was built. Multivariate logistic regression was applied to develop a clinically interpretable machine learning predictive model. The model was trained on 2016-2018 data and tested on 2019 data. A feature selection based on a univariate logistic regression first selected candidate predictors of BSI. A multivariate logistic regression with stepwise feature selection in five-fold cross-validation was applied to express the risk of BSI. A total of 5660 hospitalizations (4026 and 1634 in the training and the validation subsets, respectively) were included. Eleven predictors of BSI were identified. The performance of the model in terms of AUROC was 0.74. Based on the interquartile predicted risk score, 508 (31.1%) patients were defined as being at low risk, 776 (47.5%) at medium risk, and 350 (21.4%) at high risk of BSI. Of them, 14.2% (72/508), 30.8% (239/776), and 64% (224/350) had a BSI, respectively. The performance of the predictive model of BSI is promising. Computational infrastructure and machine learning models can help clinicians identify people at low risk for BSI, ultimately supporting an antibiotic stewardship approach.

Reduction of Aversive Learning Rates in Pavlovian Conditioning by Angiotensin II Antagonist Losartan: A Randomized Controlled Trial

BackgroundAngiotensin receptor blockade has been linked to aspects of aversive learning and memory formation and to the prevention of posttraumatic stress disorder symptom development. MethodsWe investigated the influence of the angiotensin receptor blocker losartan on aversive Pavlovian conditioning using a probabilistic learning paradigm. In a double-blind, randomized, placebo-controlled design, we tested 45 (18 female) healthy volunteers during a baseline session, after application of losartan or placebo (drug session), and during a follow-up session. During each session, participants engaged in a task in which they had to predict the probability of an electrical stimulation on every trial while the true shock contingencies switched repeatedly between phases of high and low shock threat. Computational reinforcement learning models were used to investigate learning dynamics. ResultsAcute administration of losartan significantly reduced participants’ adjustment during both low-to-high and high-to-low threat changes. This was driven by reduced aversive learning rates in the losartan group during the drug session compared with baseline. The 50-mg drug dose did not induce reduction of blood pressure or change in reaction times, ruling out a general reduction in attention and engagement. Decreased adjustment of aversive expectations was maintained at a follow-up session 24 hours later. ConclusionsThis study shows that losartan acutely reduces Pavlovian learning in aversive environments, thereby highlighting a potential role of the renin-angiotensin system in anxiety development.

Hybrid computational model depicts the contribution of non-significant lobes of human brain during the perception of emotional stimuli

Emotions are synchronizing responses of human brain while executing cognitive tasks. Earlier studies had revealed strong correlation between specific lobes of the brain to different types of emotional valence. In the current study, a comprehensive three-dimensional mapping of human brain for executing emotion specific tasks had been formulated. A hybrid computational machine learning model customized from Custom Weight Allocation Model (CWAM) and defined as Custom Rank Allocation Model (CRAM). This regression-based hybrid computational model computes the allocated tasks to different lobes of the brain during their respective executive stage. Event Related Potentials (ERP) were obtained with significant effect at P1, P2, P3, N170, N2, and N4. These ERPs were configured at Pz, Cz, F3, and T8 regions of the brain with maximal responses; while regions like Cz, C4 and F4 were also found to make effective contributions to elevate the responses of the brain, and thus these regions were configured as augmented source regions of the brain. In another circumstance of frequent –deviant - equal (FDE) presentation of the emotional stimuli, it was observed that the brain channels C3, C4, P3, P4, O1, O2, and Oz were contributing their emotional quotient to the overall response of the brain regions; whereas, the interaction effect was found presentable at O2, Oz, P3, P4, T8 and C3 regions of brain. The proposed computational model had identified the potential neural pathways during the execution of emotional task.

Application of machine learning technique for dynamic analysis of confined geomaterial subjected to vibratory load

Computer programming-based numerical programs are firmly established in geotechnical engineering, with rapid growth of finite element modeling and machine learning techniques gaining much attention both in practice and academia. This study is intended to expedite the dissemination of advanced computer applications in terms of finite element simulation and machine learning models by investigating the dynamic response of geomaterials subjected to vibratory loads. Several trial models were built to perform the experimental investigations with a vibratory shaker, signal generator, several accelerometers, a data collection system, and other ancillary devices. The implicit integration techniques in commercialized software were adopted for numerical simulations. After data collection from numerical simulation, models were chosen, trained, and assessed to produce predictions that were then used in this study. Several technologies, including the ensemble boosted tree, squared exponential Gaussian Process Regression (GPR), Matern 5/2 GPR, exponential GPR, and decision tree architectures (fine and medium), were used to forecast the displacement of confined geomaterial. The displacement-depth ratio was found rising to 80% in the frequency range of 5 to 25 Hz, suggesting a considerable change in the behavior of the geomaterial. The Matern 5/2 GPR model showed better accuracy with an R2 value of 0.99, indicating an outstanding predictive ability. The Matern 5/2 GPR and boosted tree models could help better understand the links between displacement and its distribution along the direction of load application. The outcomes of this study based on computer-aided finite element programs can be effectively implemented in machine learning to develop computer programs. In conclusion, the computational machine learning models adopted in this study offer a new insight for uncovering hidden intrinsic laws and creating new knowledge for geotechnical researchers and practitioners.