Adaptation and Validation of an Agent Model of Functional State and Performance for Individuals

Abstract

Abstract. Human performance can seriously degrade under demanding tasks. To improve performance, agents can reason about the current state of the human, and give the most appropriate and effective support. To enable this, the agent needs a model of a specific person's functional state and performance, which should be valid, as the agent might otherwise give inappropriate advice and even worsen performance. This paper concerns the adaptation of the parameters of the existing functional state model to the individual and validation of the resulting model. First, human experiments have been designed and conducted, whereby measurements related to the model have been performed. Next, this data has been used to obtain appropriate parameter settings for the model, describing the specific subject. Finally, the model, with the tailored parameter settings, has been used to predict human behavior to investigate predictive capabilities of the model. The results have been analyzed using formal verification.

Intertemporal Decision Making: A Mental Load Perspective

“…Furthermore, based on elements from an existing biologically inspired computational model for work load (cf. [26], [27], [3], [4]) a new hyperbolicexponential discounting model has been derived. This hyperbolic-exponential discounting model makes use of the concept critical power which indicates the level of power that can be provided without increasing fatigue (e.g., [8], [11], [12], [15], [19], [26], [27]).…”

“…There are two advantages of this perspective. One is that computational models are available in physiological and cognitive sciences (e.g., [3], [4], [26], [27]) that describe mental load. A second advantage is that according to this perspective value is not increasing just due to passing of time, but is explained in terms of extra work invested, in line with basic intuitions in economics, for example, behind the well-known Labour Theory of Value (cf.…”

A virtual human agent model with behaviour based on feeling exhaustion

“…A first test of the OFS-model against empirical data in a simulation-based training environment 3 has shown preliminary results which were positive; see [2]. In the validation study 31 persons participated (18 males, 13 females, of which 25 students).…”

“…In [2] it is reported how automated parameter estimation for the functional state model was used two different methods: a gradient-based approach and a simulated annealing approach based on probabilistic search. For both cases the performance level of the operator in terms of numbers of enemies shot before landing and numbers of friendly objects left unaffected was used as an evaluation criterion.…”

“…This function is essentially the probability density function of the measurement data given the parameter values. For more details, see [2]. The Simulated Annealing method used a probabilistic technique to find a parameter setting.…”