Cascading spatiotemporal pattern (CSTP) is a reflection of the underlying mechanism of natural or social processes. Identifying CSTPs provides meaningful evidence for domain experts and government agencies to forecast potential consequences of an event and plan ahead. In a natural disaster such as a hurricane, it normally results in heavy rainfall and strong winds, which will further cause floo...

The enormous popularity of Hidden Markov models (HMMs) in spatio-temporal pattern recognition is largely due to the ability to “learn” model parameters from observation sequences through the Baum-Welch and other re-estimation procedures. In this study, HMM parameters are estimated from an ensemble of models trained on individual observation sequences. The proposed methods are shown to provide s...

Examining the two motion processes is an elusive task due to the difficulty of finding a proper stimulus paradigm. A rotational Glass pattern created with a random-dot array by superimposing its rotated version on the top of it can provide such a paradigm. If we displace only its rotated part in the vertical or horizontal direction, a bistable motion occurs; local dot motion in the same directi...

A broad-bandwidth persistent spectral hole-burning organic material is used as a fast optical processor that compares an input temporal-frequency profile with a recorded reference spectral shape. This patternrecognition procedure relies on a subpicosecond temporal cross-correlation process. The size of the phaseencoding spectral interval exceeds 1 THz. The storage material, the spectral encoder...

Cross-correlation of complex, picosecond-range, temporal shapes is demonstrated experimentally with the help of a filter recorded in an organic spectral hole-burning material.

In order to exploit generalized chaotic synchronization phenomena for the solution of temporal pattern recognition problems, a chaotic dynamical system representing the class of signals that are to be recognized must be established. This system can be determined by means of identification techniques. Although the fulfillment of the chaotic condition could appear as a constraint, it is shown her...

Infectious diseases often spread as spatial epidemic outbreak waves. A number of model studies have shown that such spatial pattern formation can have important consequences for the evolution of pathogens. Here, we show that such spatial patterns can cause cyclic evolutionary dynamics in selection for the length of the infectious period. The necessary reversal in the direction of selection is e...

Although there are known algorithms for predicting observation and state sequences from HMM models there is little discussion on how to determine the contributions of the different types of HMM parameters to such predictions and consequently, temporal pattern recognition. In this note we discuss and compare a number of objective measures that provide insight into HMM performance in these terms.

With the increasing applications based on location, researches on spatiotemporal data, especially queries of spatiotemporal data have attracted a lot of attention. XML, as a standard language of information exchanging over the Web, has the ability to query spatiotemporal data. In this paper, we propose an algorithm, TwigStackSP, for matching a spatiotemporal XML query twig pattern. We represent...