Browsing by Author "Beck, Jeffrey"

Now showing 1 - 3 of 3
  • Thumbnail Image
    ItemOpen Access
    Probabilistic inferential decision-making under time pressure in rhesus macaques (Macaca mulatta)
    (Journal of Comparative Psychology) Toader, Andrew; Rao, Hrishikesh; Ryoo, Minyoung; Bohlen, Martin; Cruger, Jessi; Oh-Descher, Hanna; Ferrari, Silvia; Egner, Tobias; Beck, Jeffrey; Sommer, Marc
    Decisions often involve the consideration of multiple cues, each of which may inform selection on the basis of learned probabilities. Our ability to use probabilistic inference for decisions is bounded by uncertainty and constraints such as time pressure. Previous work showed that when humans choose between visual objects in a multiple-cue, probabilistic task, they cope with time pressure by discounting the least informative cues, an example of satisficing or “good enough” decision-making. We tested two rhesus macaques (Macaca mulatta) on a similar task to assess their capacity for probabilistic inference and satisficing in comparison with humans. On each trial, a monkey viewed two compound stimuli consisting of four cue dimensions. Each dimension (e.g., color) had two possible states (e.g., red or blue) with different probabilistic weights. Selecting the stimulus with highest total weight yielded higher odds of receiving reward. Both monkeys learned the assigned weights at high accuracy. Under time pressure, both monkeys were less accurate as a result of decreased use of cue information. One monkey adopted the same satisficing strategy used by humans, ignoring the least informative cue dimension. Both monkeys, however, exhibited a strategy not reported for humans, a “group-the-best” strategy in which the top two cues were used similarly despite their different assigned weights. The results validate macaques as an animal model of probabilistic decision-making, establishing their capacity to discriminate between objects using at least four visual dimensions simultaneously. The time pressure data suggest caution, however, in using macaques as models of human satisficing.
  • Thumbnail Image
    ItemOpen Access
    Properties of decision-making tasks govern the tradeoff between model-based and model-free learning
    (bioRxiv, 2019-08-14) Abzug, Zachary; Sommer, Marc; Beck, Jeffrey
    AbstractWhen decisions must be made between uncertain options, optimal behavior depends on accurate estimations of the likelihoods of different outcomes. The contextual factors that govern whether these estimations depend on model-free learning (tracking outcomes) vs. model-based learning (learning generative stimulus distributions) are poorly understood. We studied model-free and model-based learning using serial decision-making tasks in which subjects selected a rule and then used it to flexibly act on visual stimuli. A factorial approach defined a family of behavioral models that could integrate model-free and model-based strategies to predict rule selection trial-by-trial. Bayesian model selection demonstrated that the subjects strategies varied depending on lower-level task characteristics such as the identities of the rule options. In certain conditions, subjects integrated learned stimulus distributions and tracked reward rates to guide their behavior. The results thus identify tradeoffs between model-based and model-free decision strategies, and in some cases parallel utilization, depending on task context.
  • Thumbnail Image
    ItemOpen Access
    Variational Inference for Nonlinear Regression Using Dimension Reduced Mixtures of Generalized Linear Models with Application to Neural Data
    (2015) Subramanian, Vivek Anand

    Brain-machine interfaces (BMIs) are devices that transform neural activity into commands executed by a robotic actuator. For paraplegics who have suffered spinal cord injury and for amputees, BMIs provide an avenue to regain lost limb mobility by providing a direct connection between the brain and an actuator. One of the most important aspects of a BMI is the decoding algorithm, which interprets patterns of neural activity and issues an appropriate kinematic action. The decoding algorithm relies heavily on a neural tuning function for each neuron which describes the response of that neuron to an external stimulus or upcoming motor action. Modern BMI decoders assume a simple parametric form for this tuning function such as cosine, linear, or quadratic, and fit parameters of the chosen function to a training data set. While this may be appropriate for some neurons, tuning curves for all neurons may not all take the same parametric form; hence, performance of BMI decoding may suffer because of an inappropriate mapping from firing rate to kinematic. In this work, we develop a non-parametric model for the identification of non-linear tuning curves with arbitrary shape. We also develop an associated variational Bayesian (VB) inference scheme which provides a fast, big data-friendly method to obtain approximate posterior distributions on model parameters. We demonstrate our model's capabilities on both simulated and experimental datasets.