scholarly journals Rapid Bayesian learning in the mammalian olfactory system

2019 ◽  
Author(s):  
Naoki Hiratani ◽  
Peter E. Latham
Keyword(s):  
Olfactory System ◽  
Bayesian Learning ◽  
Experimental Studies ◽  
Bayesian Optimization ◽  
Mammalian Brain ◽  
Theoretical Understanding ◽  
Plasticity Mechanism ◽  
Neural Architecture ◽  
Learning Rules ◽  
Efficient Learning

AbstractMany experimental studies suggest that animals can rapidly learn to identify odors and predict the rewards associated with them. However, the underlying plasticity mechanism remains elusive. In particular, it is not clear how olfactory circuits achieve rapid, data efficient learning with local synaptic plasticity. Here, we formulate olfactory learning as a Bayesian optimization process, then map the learning rules into a computational model of the mammalian olfactory circuit. The model is capable of odor identification from a small number of observations, while reproducing cellular plasticity commonly observed during development. We extend the framework to reward-based learning, and show that the circuit is able to rapidly learn odor-reward association with a plausible neural architecture. These results deepen our theoretical understanding of unsupervised learning in the mammalian brain.

scholarly journals Task space adaptation via the learning of gait controllers of magnetic soft millirobots

2021 ◽  
pp. 027836492110218
Author(s):  
Sinan O. Demir ◽  
Utku Culha ◽  
Alp C. Karacakol ◽  
Abdon Pena-Francesch ◽  
Sebastian Trimpe ◽  
...  
Keyword(s):  
Human Body ◽  
Degrees Of Freedom ◽  
Bayesian Optimization ◽  
Small Scale ◽  
Soft Robots ◽  
Modeling And Control ◽  
Walking Gait ◽  
Physical Experiments ◽  
Efficient Learning ◽  
Task Environments

Untethered small-scale soft robots have promising applications in minimally invasive surgery, targeted drug delivery, and bioengineering applications as they can directly and non-invasively access confined and hard-to-reach spaces in the human body. For such potential biomedical applications, the adaptivity of the robot control is essential to ensure the continuity of the operations, as task environment conditions show dynamic variations that can alter the robot’s motion and task performance. The applicability of the conventional modeling and control methods is further limited for soft robots at the small-scale owing to their kinematics with virtually infinite degrees of freedom, inherent stochastic variability during fabrication, and changing dynamics during real-world interactions. To address the controller adaptation challenge to dynamically changing task environments, we propose using a probabilistic learning approach for a millimeter-scale magnetic walking soft robot using Bayesian optimization (BO) and Gaussian processes (GPs). Our approach provides a data-efficient learning scheme by finding the gait controller parameters while optimizing the stride length of the walking soft millirobot using a small number of physical experiments. To demonstrate the controller adaptation, we test the walking gait of the robot in task environments with different surface adhesion and roughness, and medium viscosity, which aims to represent the possible conditions for future robotic tasks inside the human body. We further utilize the transfer of the learned GP parameters among different task spaces and robots and compare their efficacy on the improvement of data-efficient controller learning.

Microemulsion Phase Behavior Observations, Thermodynamic Essentials, Mathematical Simulation

1981 ◽  
Vol 21 (06) ◽  
pp. 747-762 ◽  
Author(s):  
Karl E. Bennett ◽  
Craig H.K. Phelps ◽  
H. Ted Davis ◽  
L.E. Scriven
Keyword(s):  
Experimental Data ◽  
Phase Behavior ◽  
Experimental Studies ◽  
Mathematical Framework ◽  
Alkyl Aryl Sulfonate ◽  
Phase Volume ◽  
Displacement Efficiency ◽  
Alkyl Aryl ◽  
Theoretical Understanding ◽  
Tie Lines

Abstract The phase behavior of microemulsions of brine, hydrocarbon, alcohol, and a pure alkyl aryl sulfonate-sodium 4-(1-heptylnonyl) benzenesulfonate (SHBS or Texas 1) was investigated as a function of the concentration of salt (NaCl, MgCl2, or CaCl2), the hydrocarbon (n-alkanes, octane to hexadecane), the alcohol (butyl and amyl isomers), the concentration of surfactant, and temperature. The phase behavior mimics that of similar systems with the commercial surfactant Witco TRS 10–80. The phase volumes follow published trends, though with exceptions.A mathematical framework is presented for modeling phase behavior in a manner consistent with the thermodynamically required critical tie lines and plait point progressions from the critical endpoints. Hand's scheme for modeling binodals and Pope and Nelson's approach to modeling the evolution of the surfactant-rich third phase are extended to satisfy these requirements.An examination of model-generated progressions of ternary phase diagrams enhances understanding of the experimental data and reveals correlations of relative phase volumes (volume uptakes) with location of the mixing point (overall composition) relative to the height of the three-phase region and the locations of the critical tie lines (critical endpoints and conjugate phases). The correlations account, on thermodynamic grounds, for cases in which the surfactant is present in more than one phase or the phase volumes change discontinuously, both cases being observed in the experimental study. Introduction The phase behavior of a surfactant-based micellar formulation is one of the major factors governing the displacement efficiency of any chemical flooding process employing that formulation. Knowledge of phase behavior is, thus, important for the interpretation of laboratory core floods, the design of flooding processes, and the evaluation of field tests. Phase behavior is connected intimately with other determinants of the flooding process, such as interfacial tension and viscosity. Since the number of equilibrium phases and their volumes and appearances are easier to measure and observe than phase compositions, viscosities, and interfacial tensions, there is great interest in understanding the phase-volume/phase-property relationships. Commercial surfactants, such as Witco TRS 10-80, are sulfonates of crude or partially refined oil. While they seem to be the most economically practicable surfactants for micellar flooding, their behavior, particularly with crude oils and reservoir brines, can be difficult to interpret, the phases varying with time and from batch to batch. Phase behavior studies with a small number of components, in conjunction with a theoretical understanding of phase behavior progressions, can aid in understanding more complex behavior. In particular, one can begin to appreciate which seemingly abnormal experimental observations (e.g., surfactant present in more than one phase or a discontinuity in phase volume trends) are merely features of certain regions of any phase diagram and which are peculiar to the specific crude oil or commercial surfactant used in the study.We report here experimental studies of the phase behavior of microemulsions of a pure sulfonate surfactant (Texas 1), a single normal alkane hydrocarbon, a simple brine, and a small amount of a suitable alcohol as cosurfactant or cosolvent. The controlled variables are hydrocarbon chain length, alcohol, salinity, salt type (NaCl, MgCl2, or CaCl2), surfactant purity, surfactant concentration, and temperature. Many of these experimental data were presented earlier. SPEJ P. 747^

scholarly journals Modelling human active search in optimizing black-box functions

2020 ◽  
Vol 24 (23) ◽  
pp. 17771-17785
Author(s):  
Antonio Candelieri ◽  
Riccardo Perego ◽  
Ilaria Giordani ◽  
Andrea Ponti ◽  
Francesco Archetti
Keyword(s):  
Active Learning ◽  
Bayesian Learning ◽  
Surrogate Models ◽  
Black Box ◽  
Bayesian Optimization ◽  
Learning Approaches ◽  
Function Learning ◽  
The Subject ◽  
Human Function ◽  
Exploration Exploitation

AbstractModelling human function learning has been the subject of intense research in cognitive sciences. The topic is relevant in black-box optimization where information about the objective and/or constraints is not available and must be learned through function evaluations. In this paper, we focus on the relation between the behaviour of humans searching for the maximum and the probabilistic model used in Bayesian optimization. As surrogate models of the unknown function, both Gaussian processes and random forest have been considered: the Bayesian learning paradigm is central in the development of active learning approaches balancing exploration/exploitation in uncertain conditions towards effective generalization in large decision spaces. In this paper, we analyse experimentally how Bayesian optimization compares to humans searching for the maximum of an unknown 2D function. A set of controlled experiments with 60 subjects, using both surrogate models, confirm that Bayesian optimization provides a general model to represent individual patterns of active learning in humans.

scholarly journals A Neural Network Model Can Explain Ventriloquism Aftereffect and Its Generalization across Sound Frequencies

2013 ◽  
Vol 2013 ◽  
pp. 1-17 ◽  
Author(s):  
Elisa Magosso ◽  
Filippo Cona ◽  
Mauro Ursino
Keyword(s):  
Neural Network ◽  
Visual Stimulus ◽  
Hebbian Learning ◽  
Experimental Studies ◽  
Auditory Stimuli ◽  
Spectral Features ◽  
Auditory Neurons ◽  
Learning Rules ◽  
Perceptual Shift ◽  
Ventriloquism Aftereffect

Exposure to synchronous but spatially disparate auditory and visual stimuli produces a perceptual shift of sound location towards the visual stimulus (ventriloquism effect). After adaptation to a ventriloquism situation, enduring sound shift is observed in the absence of the visual stimulus (ventriloquism aftereffect). Experimental studies report opposing results as to aftereffect generalization across sound frequencies varying from aftereffect being confined to the frequency used during adaptation to aftereffect generalizing across some octaves. Here, we present an extension of a model of visual-auditory interaction we previously developed. The new model is able to simulate the ventriloquism effect and, via Hebbian learning rules, the ventriloquism aftereffect and can be used to investigate aftereffect generalization across frequencies. The model includes auditory neurons coding both for the spatial and spectral features of the auditory stimuli and mimicking properties of biological auditory neurons. The model suggests that different extent of aftereffect generalization across frequencies can be obtained by changing the intensity of the auditory stimulus that induces different amounts of activation in the auditory layer. The model provides a coherent theoretical framework to explain the apparently contradictory results found in the literature. Model mechanisms and hypotheses are discussed in relation to neurophysiological and psychophysical data.

scholarly journals Bayesian Inference Based on Stationary Fokker-Planck Sampling

2010 ◽  
Vol 22 (6) ◽  
pp. 1573-1596 ◽  
Author(s):  
Arturo Berrones
Keyword(s):  
Bayesian Inference ◽  
Bayesian Learning ◽  
Likelihood Estimation ◽  
Posterior Density ◽  
Step Size ◽  
Inference Models ◽  
Small Set ◽  
Efficient Learning ◽  
The Given ◽  
Fokker Planck

A novel formalism for Bayesian learning in the context of complex inference models is proposed. The method is based on the use of the stationary Fokker-Planck (SFP) approach to sample from the posterior density. Stationary Fokker-Planck sampling generalizes the Gibbs sampler algorithm for arbitrary and unknown conditional densities. By the SFP procedure, approximate analytical expressions for the conditionals and marginals of the posterior can be constructed. At each stage of SFP, the approximate conditionals are used to define a Gibbs sampling process, which is convergent to the full joint posterior. By the analytical marginals efficient learning methods in the context of artificial neural networks are outlined. Offline and incremental Bayesian inference and maximum likelihood estimation from the posterior are performed in classification and regression examples. A comparison of SFP with other Monte Carlo strategies in the general problem of sampling from arbitrary densities is also presented. It is shown that SFP is able to jump large low-probability regions without the need of a careful tuning of any step-size parameter. In fact, the SFP method requires only a small set of meaningful parameters that can be selected following clear, problem-independent guidelines. The computation cost of SFP, measured in terms of loss function evaluations, grows linearly with the given model's dimension.

scholarly journals Redundancy in synaptic connections enables neurons to learn optimally

2017 ◽  
Author(s):  
Naoki Hiratani ◽  
Tomoki Fukai
Keyword(s):  
Synaptic Plasticity ◽  
Particle Filtering ◽  
Neuron Model ◽  
Experimental Studies ◽  
Dendritic Tree ◽  
Spike Timing ◽  
Mammalian Brain ◽  
Synaptic Connections ◽  
Synaptic Organization ◽  
Functional Benefit

AbstractRecent experimental studies suggest that, in cortical microcircuits of the mammalian brain, the majority of neuron-to-neuron connections are realized by multiple synapses. However, it is not known whether such redundant synaptic connections provide any functional benefit. Here, we show that redundant synaptic connections enable near-optimal learning in cooperation with synaptic rewiring. By constructing a simple dendritic neuron model, we demonstrate that with multisynaptic connections, synaptic plasticity approximates a sample-based Bayesian filtering algorithm known as particle filtering, and wiring plasticity implements its resampling process. Applying the proposed framework to a detailed single neuron model, we show that the model accounts for many experimental observations, including the dendritic position dependence of spike-timing-dependent plasticity, and the functional synaptic organization on the dendritic tree based on the stimulus selectivity of presynaptic neurons. Our study provides a novel conceptual framework for synaptic plasticity and rewiring.

scholarly journals Computational Insights Into Avidity of Polymeric Multivalent Binders

2019 ◽  
Author(s):  
E Zumbro ◽  
J Witten ◽  
A Alexander-Katz
Keyword(s):  
Brownian Dynamics ◽  
Structural Parameters ◽  
Experimental Studies ◽  
Degree Of Polymerization ◽  
Design Strategies ◽  
Theoretical Understanding ◽  
Biological Design ◽  
Multivalent Binding ◽  
Dynamics Simulations ◽  
Binding Avidity

AbstractMultivalent binding interactions are commonly found throughout biology to enhance weak monovalent binding such as between glycoligands and protein receptors. Designing multivalent polymers to bind to viruses and toxic proteins is a promising avenue for inhibiting their attachment and subsequent infection of cells. Several studies have focused on oligomeric multivalent inhibitors and how changing parameters such as ligand shape and size, and linker length and flexibility affect binding. However, experimental studies of how larger structural parameters of multivalent polymers such as degree of polymerization affect binding avidity to targets have mixed results with some finding an improvement with longer polymers and some finding no effect. Here, we use Brownian dynamics simulations to provide a theoretical understanding of how degree of polymerization affects the binding avidity of multivalent polymers. We show that longer polymers increase binding avidity to multivalent targets, but reach a limit in binding avidity at high degrees of polymerization. We also show that when interacting with multiple targets simultaneously, longer polymers are able to use inter-target interactions to promote clustering and improve binding efficiency. We expect our results to narrow the design space for optimizing the structure and effectiveness of multivalent inhibitors, as well as be useful to understand biological design strategies for multivalent binding.

scholarly journals Sparse representation for damage identification of structural systems

2020 ◽  
pp. 147592172092697
Author(s):  
Zhao Chen ◽  
Hao Sun
Keyword(s):  
Sparse Representation ◽  
Structural Damage ◽  
Bayesian Learning ◽  
Damage Identification ◽  
Model Updating ◽  
Computational Cost ◽  
Bayesian Optimization ◽  
Structural Systems ◽  
Least Squares Regression ◽  
Shear Type

Identifying damage of structural systems is typically characterized as an inverse problem which might be ill-conditioned due to aleatory and epistemic uncertainties induced by measurement noise and modeling error. Sparse representation can be used to perform inverse analysis for the case of sparse damage. In this article, we propose a novel two-stage sensitivity analysis–based framework for both model updating and sparse damage identification. Specifically, an [Formula: see text] Bayesian learning method is first developed for updating the intact model and uncertainty quantification so as to set forward a baseline for damage detection. A sparse representation pipeline built on a quasi-[Formula: see text] method, for example, sequential threshold least squares regression, is then presented for damage localization and quantification. In addition, Bayesian optimization together with cross-validation is developed to heuristically learn hyperparameters from data, which saves the computational cost of hyperparameter tuning and produces more reliable identification result. The proposed framework is verified by three examples, including a 10-story shear-type building, a complex truss structure, and a shake-table test of an eight-story steel frame. Results show that the proposed approach is capable of both localizing and quantifying structural damage with high accuracy.

Collaborative Bayesian Image Annotation and Retrieval

2011 ◽  
pp. 146-169
Author(s):  
Rui Zhang ◽  
Ling Guan
Keyword(s):  
Image Retrieval ◽  
Life Experience ◽  
Image Annotation ◽  
Contextual Information ◽  
Experimental Studies ◽  
Real Life ◽  
Learning Techniques ◽  
Semantic Concepts ◽  
Efficient Learning ◽  
Classification Procedures

With nearly twenty years of intensive study on the content-based image retrieval and annotation, the topic still remains difficult. By and large, the essential challenge lies in the limitation of using low-level visual features to characterize the semantic information of images, commonly known as the semantic gap. To bridge this gap, various approaches have been proposed based on the incorporation of human knowledge and textual information as well as the learning techniques utilizing the information of different modalities. At the same time, contextual information which represents the relationship between different real world/conceptual entities has shown its significance with respect to recognition tasks not only through real life experience but also scientific studies. In this chapter, the authors first review the state of the art of the existing works on image annotation and retrieval. Moreover, a general Bayesian framework which integrates content and contextual information and its application to both image annotation and retrieval are elaborated. The contextual information is considered as the statistical relationship between different images and different semantic concepts for image retrieval and annotation, respectively. The framework has efficient learning and classification procedures and the effectiveness is evaluated based on experimental studies, which demonstrate its advantage over both content-based and context-based approaches.

Sign in / Sign up

Export Citation Format

Share Document