A regularization method for linking brain and behavior.

Kang, Inhan; Yi, Woojong; Turner, B. E.

doi:10.1037/met0000387

Cited by 11 publications

(12 citation statements)

References 57 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We found that there were no strong relationships between the parameters both within-component and between the neural and behavioral parameters. The factor loading estimates in our results do not resemble the loadings found in Turner, Wang, and Merkle (2017) and Kang et al (2021), where we can see the extent of the potential bias of the loadings estimates resulting from data pooling, which also fails to account for individual differences in the data. Additionally, the true relationship between the neural and behavioral parameters is likely obscured because of the low number of trials per participant.…”

Section: Application 3: Neural and Behavioural Datacontrasting

confidence: 99%

“…Evidently, the method shares many parallels with joint-modelling work by Turner, Wang, and Merkle (2017), however, it extends on this approach by offering a more flexible, hierarchical sampling method, which draws strengths from state of the art Bayesian estimation techniques (namely PMwG; Gunawan et al, 2020). Further, methods by Kang et al (2021) that build on the factor analysis methods of Turner, Wang, and Merkle (2017) by including Lasso priors, could also extend to the current methods. Whilst the initial methods of Turner, Wang, and Merkle (2017) are useful for estimating factor loadings of neural activity related to the NDDM (Turner, Van Maanen, & Forstmann, 2015), PMwG-FA allows model estimation with a factor analysis Gibbs step, in a Bayesian hierarchical framework, to be applied to any statistical model (where estimating the likelihood is possible).…”

Section: Discussionmentioning

confidence: 99%

“…Whilst the initial methods of Turner, Wang, and Merkle (2017) are useful for estimating factor loadings of neural activity related to the NDDM (Turner, Van Maanen, & Forstmann, 2015), PMwG-FA allows model estimation with a factor analysis Gibbs step, in a Bayesian hierarchical framework, to be applied to any statistical model (where estimating the likelihood is possible). This means that in contrast to Turner, Wang, and Merkle (2017) and Kang et al (2021), highly specific assumptions regarding the factors are not necessary and factor analysis can be conducted using exploratory methods. Further, PMwG-FA allows a fully Bayesian hierarchical method of estimation, with subject level and group level models estimated that are reciprocally related, hence respecting the individual differences present in the data.…”

Section: Discussionmentioning

confidence: 99%

“…However, this post-hoc factor analysis approach is likely to suffer from the aforementioned attenuation, which tends to accompany two-step analysis of correlations (Matzke et al, 2017;. Similarly, the method shown by Turner, Wang, and Merkle (2017) and Kang, Yi, and Turner (2021) uses simplifications of the factors and model parameters estimated that make very strong assumptions about the underlying factors that may exist in the data, specifically that the assumptions of the cognitive model underpin the architecture of the linking function. This means an explicit assumption is made that the cognitive model drives both the neural and behavioural data, which in turn means that this method is difficult to generalise to alternate model-based ques-tions.…”

Section: Factor Analysis In Joint Modellingmentioning

confidence: 99%

See 3 more Smart Citations

Using group level factor models to resolve high dimensionality in model-based sampling

Innes¹,

Stevenson²,

Gronau³

et al. 2022

Preprint

View full text Add to dashboard Cite

Joint modelling of behaviour and neural activation poses the potential to provide significant advances in linking brain and behaviour. However, methods of joint modelling have been limited by difficulties in estimation, often due to high dimensionality and simultaneous estimation challenges. In the current article, we propose a method of model estimation which draws on current state-of-the-art Bayesian hierarchical modelling techniques and uses factor analysis as a means of dimensionality reduction to provide further information on which to make inference. The method uses a particle metropolis within Gibbs sampler (PMwG; Gunawan, Hawkins, Tran, Kohn, & Brown, 2020), where the factor structure is estimated within the Gibbs step for the group level. We show the significant dimensionality reduction gained by factor analysis in the Gibbs step of the PMwG, evidence for parameter recovery, a variety of factor loading constraints which can be used for different purposes and research questions, as well as two applications of the method to previously analysed data. This method represents a flexible and usable approach with interpretable outcomes, which relies on data driven analysis as opposed to hypothesis driven methods often used in joint modelling. Although we focus on joint modelling methods, this model based estimation approach could be used for any high dimensional modelling problem. We provide open source code and accompanying tutorial documentation to make the method accessible to any researchers.

show abstract

Section: Application 3: Neural and Behavioural Datacontrasting

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Section: Factor Analysis In Joint Modellingmentioning

confidence: 99%

See 2 more Smart Citations

Using group level factor models to resolve high dimensionality in model-based sampling

Innes¹,

Stevenson²,

Gronau³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…These neuro-cognitive models utilize trial-or individual-level signatures of EEG or magnetoencephalography (MEG) signals to assess, constrain, replace, or even add cognitive parameters in models (Nunez et al, 2017(Nunez et al, , 2022. Moreover, Turner et al have proposed many approaches for directly or indirectly relating neural data to a cognitive model; for instance, the field of model-based cognitive neuroscience has used BOLD responses and EEG waveforms simultaneously and separately to predict and constrain cognitive parameters and behavioral data (Turner et al, 2013(Turner et al, , 2016(Turner et al, , 2019Bahg et al, 2020;Kang et al, 2021). Also, Nunez et al have introduced some neuro-cognitive models to study the effect of selective attention, the role of visual encoding time (VET), and the relationship of readiness potentials (RPs) in motor cortical areas to evidence accumulation during perceptual decision-making tasks (Nunez et al, 2017(Nunez et al, , 2019Lui et al, 2021).…”

Section: Introductionmentioning

confidence: 99%

A general integrative neurocognitive modeling framework to jointly describe EEG and decision-making on single trials

Ghaderi-Kangavari¹,

Rad²,

Nunez³

2022

Preprint

View full text Add to dashboard Cite

Despite advances in techniques for exploring reciprocity in brain-behavior relations, few studies focus on building neurocognitive models that describe both human EEG and behavioral modalities at the single-trial level. Here, we introduce a new integrative joint modeling framework for the simultaneous description of single-trial EEG measures and cognitive modeling parameters of decision making. The new framework can be utilized for the evaluation of research questions as well as the prediction of both data types concurrently. In the introduced joint models, we formalized how single-trial N200 latencies and Centro-parietal positivities (CPP) are predicted by changing single-trial parameters of various drift-diffusion models (DDMs). These models do not have clear closed-form likelihoods and are not easy to fit using Markov chain Monte Carlo (MCMC) methods because nuisance parameters on single trials are shared in both behavior and neural activity. We trained deep neural networks to learn the Bayesian posterior distributions of unobserved neurocognitive parameters based on model simulations. We then used parameter recovery assessment and model misspecification to ascertain how robustly the models’ parameters can be estimated. Moreover, we fit the models to three different real datasets to test their applicability. Our results show that the single-trial integrative models can recover their latent parameters. Finally, we provide some pieces of evidence that single-trial integrative joint models are superior to traditional integrative models. The current single-trial paradigm and the likelihood-free approach for parameter recovery can inspire scientists and modelers to conveniently develop new neuro-cognitive models for other neural measures and to evaluate them appropriately.

show abstract

A General Integrative Neurocognitive Modeling Framework to Jointly Describe EEG and Decision-making on Single Trials

2023

View full text Add to dashboard Cite

Despite advances in techniques for exploring reciprocity in brain-behavior relations, few studies focus on building neurocognitive models that describe both human EEG and behavioral modalities at the single-trial level. Here, we introduce a new integrative joint modeling framework for the simultaneous description of single-trial EEG measures and cognitive modeling parameters of decision-making. As specific examples, we formalized how single-trial N200 latencies and centro-parietal positivities (CPPs) are predicted by changing single-trial parameters of various drift-diffusion models (DDMs). We trained deep neural networks to learn Bayesian posterior distributions of unobserved neurocognitive parameters based on model simulations. These models do not have closed-form likelihoods and are not easy to fit using Markov chain Monte Carlo (MCMC) methods because nuisance parameters on single trials are shared in both behavior and neural activity. We then used parameter recovery assessment and model misspecification to ascertain how robustly the models’ parameters can be estimated. Moreover, we fit the models to three different real datasets to test their applicability. Finally, we provide some evidence that single-trial integrative joint models are superior to traditional integrative models. The current single-trial paradigm and the simulation-based (likelihood-free) approach for parameter recovery can inspire scientists and modelers to conveniently develop new neurocognitive models for other neural measures and to evaluate them appropriately.

show abstract

A regularization method for linking brain and behavior.

Cited by 11 publications

References 57 publications

Using group level factor models to resolve high dimensionality in model-based sampling

Using group level factor models to resolve high dimensionality in model-based sampling

A general integrative neurocognitive modeling framework to jointly describe EEG and decision-making on single trials

A General Integrative Neurocognitive Modeling Framework to Jointly Describe EEG and Decision-making on Single Trials

Contact Info

Product

Resources

About