T Tambouratzis | PubMed Health

PubMed2001

String matching artificial neural networks.

Three artificial neural networks (ANNs) are proposed for solving a variety of on- and off-line string matching problems. The ANN structure employed as the building block of these ANNs is derived from the harmony theory (HT) ANN, whereby the resulting string matching ANNs are characterized by fast match-mismatch decisions, low computational complexity, and activation values of the ANN output nodes that can be used as indicators of substitution, insertion (addition) and deletion spelling errors.

Artificial Intelligence↗

PubMed1993

The relative effectiveness of serial and parallel viewing of line drawings depicting possible and impossible objects.

In a series of experiments, subjects were asked to make judgements concerning the three-dimensional constructibility of line drawings depicting possible and impossible objects. A spectrum of objects was employed in which complexity as well as, for impossible objects, the cause and saliency of the contradiction in three-dimensional structure varied widely. The line drawings were presented under varying viewing conditions and exposure times. It was found that line drawings of possible objects were more often correctly identified than those of impossible ones. Parallel (simultaneous) viewing was more efficient than serial viewing (in which a line drawing moved behind a narrow stationary aperture). The orientation of the aperture did not cause differences in the subjects' performance. Line-drawing complexity and contradiction in three-dimensional structure were not found to be significant for accurate recognition. Finally, no consistent effect of exposure duration on performance could be determined in the range 60-1000 ms.

Female↗