Intervention and Causality: Forecasting Traffic Flows Using a Dynamic Bayesian Network

Hyperscanning fNIRS data analysis using multiregression dynamic models: an illustration in a violin duo

Introduction

Interpersonal neural synchronization (INS) demands a greater understanding of a brain's influence on others. Therefore, brain synchronization is an even more complex system than intrasubject brain connectivity and must be investigated. There is a need to develop novel methods for statistical inference in this context.

Methods

In this study, motivated by the analysis of fNIRS hyperscanning data, which measure the activity of multiple brains simultaneously, we propose a two-step network estimation: Tabu search local method and global maximization in the selected subgroup [partial conditional directed acyclic graph (DAG) + multiregression dynamic model]. We illustrate this approach in a dataset of two individuals who are playing the violin together.

Results

This study contributes new tools to the social neuroscience field, which may provide new perspectives about intersubject interactions. Our proposed approach estimates the best probabilistic network representation, in addition to providing access to the time-varying parameters, which may be helpful in understanding the brain-to-brain association of these two players.

Discussion

The illustration of the violin duo highlights the time-evolving changes in the brain activation of an individual influencing the other one through a data-driven analysis. We confirmed that one player was leading the other given the ROI causal relation toward the other player.

FCG-ASpredictor: An Approach for the Prediction of Average Speed of Road Segments with Floating Car GPS Data

The average speed (AS) of a road segment is an important factor for predicting traffic congestion, because the accuracy of AS can directly affect the implementation of traffic management. The traffic environment, spatiotemporal information, and the dynamic interaction between these two factors impact the predictive accuracy of AS in the existing literature, and floating car data comprehensively reflect the operation of urban road vehicles. In this paper, we proposed a novel road segment AS predictive model, which is based on floating car data. First, the impact of historical AS, weather, and date attributes on AS prediction has been analyzed. Then, through spatiotemporal correlations calculation based on the data from Global Positioning System (GPS), the predictive method utilizes the recursive least squares method to fuse the historical AS with other factors (such as weather, date attributes, etc.) and adopts an extended Kalman filter algorithm to accurately predict the AS of the target segment. Finally, we applied our approach on the traffic congestion prediction on four road segments in Chengdu, China. The results showed that the proposed predictive model is highly feasible and accurate.

A group analysis using the Multiregression Dynamic Models for fMRI networked time series

Connectivity studies of the brain are usually based on functional Magnetic Resonance Imaging (fMRI) experiments involving many subjects. These studies need to take into account not only the interaction between areas of a single brain but also the differences amongst those subjects. In this paper we develop a methodology called the group-structure (GS) approach that models possible heterogeneity between subjects and searches for distinct homogeneous sub-groups according to some measure that reflects the connectivity maps. We suggest a GS method that uses a novel distance based on a model selection measure, the Bayes factor. We then develop a new class of Multiregression Dynamic Models to estimate individual networks whilst acknowledging a GS type dependence structure across subjects. We compare the efficacy of this methodology to three other methods, virtual-typical-subject (VTS), individual-structure (IS) and common-structure (CS), used to infer a group network using both synthetic and real fMRI data. We find that the GS approach provides results that are both more consistent with the data and more flexible in their interpretative power than its competitors. In addition, we present two methods, the Individual Estimation of Multiple Networks (IEMN) and the Marginal Estimation of Multiple Networks (MEMN), generated from the GS approach and used to estimate all types of networks informed by an experiment —individual, homogeneous subgroups and group networks. These methods are then compared both from a theoretical perspective and in practice using real fMRI data.

•

The group-structure (GS) approach models possible heterogeneity between subjects.

•

The search for distinct homogeneous groups is based on the connectivity maps.

•

All four methods for group network were compared using synthetic and real fMRI data

•

IEMN and MEMN estimate individual, homogeneous subgroups and group networks.

Toward a multisubject analysis of neural connectivity

Directed acyclic graphs (DAGs) and associated probability models are widely used to model neural connectivity and communication channels. In many experiments, data are collected from multiple subjects whose connectivities may differ but are likely to share many features. In such circumstances, it is natural to leverage similarity among subjects to improve statistical efficiency. The first exact algorithm for estimation of multiple related DAGs was recently proposed by Oates, Smith, Mukherjee, and Cussens ( 2014 ). In this letter we present examples and discuss implications of the methodology as applied to the analysis of fMRI data from a multisubject experiment. Elicitation of tuning parameters requires care, and we illustrate how this may proceed retrospectively based on technical replicate data. In addition to joint learning of subject-specific connectivity, we allow for heterogeneous collections of subjects and simultaneously estimate relationships between the subjects themselves. This letter aims to highlight the potential for exact estimation in the multisubject setting.