A micro-genesis account of longer-form reinforcement learning in structured and unstructured environments.
Ontology highlight
ABSTRACT:
SUBMITTER: Dyson BJ
PROVIDER: S-EPMC8222288 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:Reinforcement learning involves updating estimates of the value of states and actions on the basis of experience. Previous work has shown that in humans, reinforcement learning exhibits a confirmatory bias: when the value of a chosen option is being updated, estimates are revised more radically following positive than negative reward prediction errors, but the converse is observed when updating the unchosen option value estimate. Here, we simulate performance on a multi-arm bandit task to examine the consequences of a confirmatory bias for reward harvesting. We report a paradoxical finding: that confirmatory biases allow the agent to maximize reward relative to an unbiased updating rule. This principle holds over a wide range of experimental settings and is most influential when decisions are corrupted by noise. We show that this occurs because on average, confirmatory biases lead to overestimating the value of more valuable bandits and underestimating the value of less valuable bandits, rendering decisions overall more robust in the face of noise. Our results show how apparently suboptimal learning rules can in fact be reward maximizing if decisions are made with finite computational precision.
| S-EPMC7612695 | biostudies-literature
Project description:The Value Learning Task (VLT; e.g., Raymond & O'Brien, 2009) is widely used to investigate how acquired value impacts how we perceive and process stimuli. The task consists of a series of trials in which participants attempt to maximize accumulated winnings as they make choices from a pair of presented images associated with probabilistic win, loss, or no-change outcomes. The probabilities and outcomes are initially unknown to the participant and thus the task involves decision making and learning under uncertainty. Despite the symmetric outcome structure for win and loss pairs, people learn win associations better than loss associations (Lin, Cabrera-Haro, & Reuter-Lorenz, 2020). This learning asymmetry could lead to differences when the stimuli are probed in subsequent tasks, compromising inferences about how acquired value affects downstream processing. We investigate the nature of the asymmetry using a standard error-driven reinforcement learning model with a softmax choice rule. Despite having no special role for valence, the model yields the learning asymmetry observed in human behavior, whether the model parameters are set to maximize empirical fit, or task payoff. The asymmetry arises from an interaction between a neutral initial value estimate and a choice policy that exploits while exploring, leading to more poorly discriminated value estimates for loss stimuli. We also show how differences in estimated individual learning rates help to explain individual differences in the observed win-loss asymmetries, and how the final value estimates produced by the model provide a simple account of a post-learning explicit value categorization task.
| S-EPMC10390629 | biostudies-literature
Project description:Humans have the fascinating ability to achieve goals in a complex and constantly changing world, still surpassing modern machine-learning algorithms in terms of flexibility and learning speed. It is generally accepted that a crucial factor for this ability is the use of abstract, hierarchical representations, which employ structure in the environment to guide learning and decision making. Nevertheless, how we create and use these hierarchical representations is poorly understood. This study presents evidence that human behavior can be characterized as hierarchical reinforcement learning (RL). We designed an experiment to test specific predictions of hierarchical RL using a series of subtasks in the realm of context-based learning and observed several behavioral markers of hierarchical RL, such as asymmetric switch costs between changes in higher-level versus lower-level features, faster learning in higher-valued compared to lower-valued contexts, and preference for higher-valued compared to lower-valued contexts. We replicated these results across three independent samples. We simulated three models-a classic RL, a hierarchical RL, and a hierarchical Bayesian model-and compared their behavior to human results. While the flat RL model captured some aspects of participants' sensitivity to outcome values, and the hierarchical Bayesian model captured some markers of transfer, only hierarchical RL accounted for all patterns observed in human behavior. This work shows that hierarchical RL, a biologically inspired and computationally simple algorithm, can capture human behavior in complex, hierarchical environments and opens the avenue for future research in this field.
| S-EPMC7703642 | biostudies-literature
Project description:We recently showed that improved perceptual performance on a visual motion direction-discrimination task corresponds to changes in how an unmodified sensory representation in the brain is interpreted to form a decision that guides behavior. Here we found that these changes can be accounted for using a reinforcement-learning rule to shape functional connectivity between the sensory and decision neurons. We modeled performance on the basis of the readout of simulated responses of direction-selective sensory neurons in the middle temporal area (MT) of monkey cortex. A reward prediction error guided changes in connections between these sensory neurons and the decision process, first establishing the association between motion direction and response direction, and then gradually improving perceptual sensitivity by selectively strengthening the connections from the most sensitive neurons in the sensory population. The results suggest a common, feedback-driven mechanism for some forms of associative and perceptual learning.
| S-EPMC2674144 | biostudies-literature
Project description:Humans and animals are endowed with a large number of effectors. Although this enables great behavioral flexibility, it presents an equally formidable reinforcement learning problem of discovering which actions are most valuable because of the high dimensionality of the action space. An unresolved question is how neural systems for reinforcement learning-such as prediction error signals for action valuation associated with dopamine and the striatum-can cope with this "curse of dimensionality." We propose a reinforcement learning framework that allows for learned action valuations to be decomposed into effector-specific components when appropriate to a task, and test it by studying to what extent human behavior and blood oxygen level-dependent (BOLD) activity can exploit such a decomposition in a multieffector choice task. Subjects made simultaneous decisions with their left and right hands and received separate reward feedback for each hand movement. We found that choice behavior was better described by a learning model that decomposed the values of bimanual movements into separate values for each effector, rather than a traditional model that treated the bimanual actions as unitary with a single value. A decomposition of value into effector-specific components was also observed in value-related BOLD signaling, in the form of lateralized biases in striatal correlates of prediction error and anticipatory value correlates in the intraparietal sulcus. These results suggest that the human brain can use decomposed value representations to "divide and conquer" reinforcement learning over high-dimensional action spaces.
| S-EPMC2796632 | biostudies-literature
Project description:BackgroundThe broad adoption of electronic health records (EHRs) provides great opportunities to conduct health care research and solve various clinical problems in medicine. With recent advances and success, methods based on machine learning and deep learning have become increasingly popular in medical informatics. However, while many research studies utilize temporal structured data on predictive modeling, they typically neglect potentially valuable information in unstructured clinical notes. Integrating heterogeneous data types across EHRs through deep learning techniques may help improve the performance of prediction models.MethodsIn this research, we proposed 2 general-purpose multi-modal neural network architectures to enhance patient representation learning by combining sequential unstructured notes with structured data. The proposed fusion models leverage document embeddings for the representation of long clinical note documents and either convolutional neural network or long short-term memory networks to model the sequential clinical notes and temporal signals, and one-hot encoding for static information representation. The concatenated representation is the final patient representation which is used to make predictions.ResultsWe evaluate the performance of proposed models on 3 risk prediction tasks (i.e. in-hospital mortality, 30-day hospital readmission, and long length of stay prediction) using derived data from the publicly available Medical Information Mart for Intensive Care III dataset. Our results show that by combining unstructured clinical notes with structured data, the proposed models outperform other models that utilize either unstructured notes or structured data only.ConclusionsThe proposed fusion models learn better patient representation by combining structured and unstructured data. Integrating heterogeneous data types across EHRs helps improve the performance of prediction models and reduce errors.
| S-EPMC7596962 | biostudies-literature
Project description:Little is known about the relationship between attention and learning during decision making. Using eye tracking and multivariate pattern analysis of fMRI data, we measured participants' dimensional attention as they performed a trial-and-error learning task in which only one of three stimulus dimensions was relevant for reward at any given time. Analysis of participants' choices revealed that attention biased both value computation during choice and value update during learning. Value signals in the ventromedial prefrontal cortex and prediction errors in the striatum were similarly biased by attention. In turn, participants' focus of attention was dynamically modulated by ongoing learning. Attentional switches across dimensions correlated with activity in a frontoparietal attention network, which showed enhanced connectivity with the ventromedial prefrontal cortex between switches. Our results suggest a bidirectional interaction between attention and learning: attention constrains learning to relevant dimensions of the environment, while we learn what to attend to via trial and error.
| S-EPMC5287409 | biostudies-literature
Project description:Independent reinforcement learning algorithms have no theoretical guarantees for finding the best policy in multi-agent settings. However, in practice, prior works have reported good performance with independent algorithms in some domains and bad performance in others. Moreover, a comprehensive study of the strengths and weaknesses of independent algorithms is lacking in the literature. In this paper, we carry out an empirical comparison of the performance of independent algorithms on seven PettingZoo environments that span the three main categories of multi-agent environments, i.e., cooperative, competitive, and mixed. For the cooperative setting, we show that independent algorithms can perform on par with multi-agent algorithms in fully-observable environments, while adding recurrence improves the learning of independent algorithms in partially-observable environments. In the competitive setting, independent algorithms can perform on par or better than multi-agent algorithms, even in more challenging environments. We also show that agents trained via independent algorithms learn to perform well individually, but fail to learn to cooperate with allies and compete with enemies in mixed environments.
| S-EPMC9530713 | biostudies-literature
Project description:Enormous amounts of biomedical data have been and are being produced by investigators all over the world. However, one crucial and limiting factor in data reuse is accurate, structured and complete description of the data or data about the data-defined as metadata. We propose a framework to predict structured metadata terms from unstructured metadata for improving quality and quantity of metadata, using the Gene Expression Omnibus (GEO) microarray database. Our framework consists of classifiers trained using term frequency-inverse document frequency (TF-IDF) features and a second approach based on topics modeled using a Latent Dirichlet Allocation model (LDA) to reduce the dimensionality of the unstructured data. Our results on the GEO database show that structured metadata terms can be the most accurately predicted using the TF-IDF approach followed by LDA both outperforming the majority vote baseline. While some accuracy is lost by the dimensionality reduction of LDA, the difference is small for elements with few possible values, and there is a large improvement over the majority classifier baseline. Overall this is a promising approach for metadata prediction that is likely to be applicable to other datasets and has implications for researchers interested in biomedical metadata curation and metadata prediction.
| S-EPMC4892825 | biostudies-other
Project description:The objective of autonomous robotic additive manufacturing for construction in the architectural scale is currently being investigated in parts both within the research communities of computational design and robotic fabrication (CDRF) and deep reinforcement learning (DRL) in robotics. The presented study summarizes the relevant state of the art in both research areas and lays out how their respective accomplishments can be combined to achieve higher degrees of autonomy in robotic construction within the Architecture, Engineering and Construction (AEC) industry. A distributed control and communication infrastructure for agent training and task execution is presented, that leverages the potentials of combining tools, standards and algorithms of both fields. It is geared towards industrial CDRF applications. Using this framework, a robotic agent is trained to autonomously plan and build structures using two model-free DRL algorithms (TD3, SAC) in two case studies: robotic block stacking and sensor-adaptive 3D printing. The first case study serves to demonstrate the general applicability of computational design environments for DRL training and the comparative learning success of the utilized algorithms. Case study two highlights the benefit of our setup in terms of tool path planning, geometric state reconstruction, the incorporation of fabrication constraints and action evaluation as part of the training and execution process through parametric modeling routines. The study benefits from highly efficient geometry compression based on convolutional autoencoders (CAE) and signed distance fields (SDF), real-time physics simulation in CAD, industry-grade hardware control and distinct action complementation through geometric scripting. Most of the developed code is provided open source. Supplementary Information The online version contains supplementary material available at 10.1007/s41693-022-00069-0.
| S-EPMC9125977 | biostudies-literature