Title :
Vectorial phase-space analysis for detecting dynamical interactions in firing patterns of biological neural networks
Author_Institution :
Div. of Neurosci., Bayor Coll. of Med., Houston, TX, USA
Abstract :
A vectorial statistical phase-space analysis is introduced to detect temporally correlated firing patterns in a network of n neurons. The cross-interval vector is used to establish temporal correlation between firing intervals between neurons. The resultant vectorial sum computed from these individual cross-interval vectors establishes a statistical average measure of the cross correlation among all n neurons. Thus, an n-tuple correlation among all n neurons in the network can be computed. The normalized resultant vectors not only capture an O(n3) combinatorial correlation but also reduce the combinatorics to an O (n) vectorial statistic. This vectorial phase-space analysis provides a description of the temporal correlation relationships among different neurons from the trajectories of the cross-interval vectors
Keywords :
combinatorial mathematics; computational complexity; neural nets; neurophysiology; phase space methods; statistical analysis; biological neural networks; combinatorial correlation; complexity; cross-interval vector; cross-interval vector trajectories; dynamical interaction detection; temporal correlation; temporally correlated firing patterns; vectorial statistical phase-space analysis; Biological neural networks; Computer networks; Educational institutions; Electronic mail; Intelligent networks; Neurons; Neuroscience; Pattern analysis; Phase detection; Statistics;
Conference_Titel :
Neural Networks, 1992. IJCNN., International Joint Conference on
Conference_Location :
Baltimore, MD
Print_ISBN :
0-7803-0559-0
DOI :
10.1109/IJCNN.1992.227184