Abstract: 

This article reports the results of research on modeling activity pattern measurement and prediction. The theoretical background of the proposed model is the time-geographic notion of interdependency and sequential connectivity that reflects the choreography of everyday life. The model employs and extends biological sequential alignment methods to address the multidimensional nature of activity patterns. The model predicts activity patterns as a result of a set of decision heuristics and activity utility functions. The measurement and prediction models are complementary: the measurement model can be used to identify segments of homogeneous activity patterns so that the prediction model can be estimated for each segment separately. The measurement model could possibly also be used to measure the goodness-of-fit between predicted and observed activity patterns in a parameter estimation process. The face validity of the proposed models was assessed using activity diary data, the results of which support the suggested approach.