Search results

In this chapter we highlight a neurodynamic approach that is showing promise as a quantitative measure of team performance.

During teamwork the rapid electroencephalographic (EEG) oscillations that emerge on the scalp were transformed into symbolic data streams which provided historical details at a second-by-second resolution of how the team perceived the evolving task and how they adjusted their dynamics to compensate for, and anticipate new task challenges. Key to this approach are the different strategies that can be used to reduce the data dimensionality, including compression, abstraction and taking advantage of the natural redundancy in biologic signals.

The framework emerging is that teams continually enter and leave organizational neurodynamic partnerships with each other, so-called metastable states, depending on the evolving task, with higher level dynamics arising from mechanisms that naturally integrate over faster microscopic dynamics.

The development of quantitative measures of the momentary dynamics of teams is anticipated to significantly influence how teams are assembled, trained, and supported. The availability of such measures will enable objective comparisons to be made across teams, training protocols, and training sites. They will lead to better understandings of how expertise is developed and how training can be modified to accelerate the path toward expertise.

The innovation of this study is the potential it raises for developing globally applicable quantitative models of team dynamics that will allow comparisons to be made across teams, tasks, and training protocols.

This chapter describes a neurodynamic modeling approach which may be useful for dynamically assessing teamwork in healthcare and military situations. It begins with a description of electroencephalographic (EEG) signal acquisition and the transformation of the physical units of EEG signals into quantities of information. This transformation provides quantitative, dynamic, and generalizable neurodynamic models that are directly comparable across teams, tasks, training protocols, and team experience levels using the same measurement scale, bits of information. These bits of information can be further used to dynamically guide team performance or to provide after-action feedback that is linked to task events and team actions.

These ideas are instantiated and expanded in the second section of the chapter by showing how these data abstractions, compressions, and transformations take advantage of the natural information redundancy in biologic signals to substantially reduce the number of data dimensions, making the incorporation of neurodynamic feedback into Intelligent Tutoring Systems (ITSs) achievable.

This chapter introduces a risk control framework on credit card fraud instead of providing a solely binary classifier model. The anomaly detection approach is adopted to identify fraud events as the outliers of the reconstruction error of a trained autoencoder (AE). The trained AE shows fitness and robustness on the normal transactions and heterogeneous behavior on fraud activities. The cost of false-positive normal transactions is controlled, and the loss of false-negative frauds can be evaluated by the thresholds from the percentiles of reconstruction error of trained AE on normal transactions. To align the risk assessment of the economic and financial situation, the risk manager can adjust the threshold to meet the risk control requirements. Using the 95th percentile as the threshold, the rate of wrongly detecting normal transactions is controlled at 5% and the true positive rate is 86%. For the 99th percentile threshold, the well-controlled false positive rate is around 1% and 83% for the truly detecting fraud activities. The performance of a false positive rate and the true positive rate is competitive with other supervised learning algorithms.