Abstract
A neuroelectric signal recognition program written for a DEC PDP-12 digital computer is described. Unit selections are made on the basis of the minimal mean square amplitude differences between the incoming displayed signal and previous “template unit” selections. These templates are continuously being replaced by the incoming “units of best fit”, thus ensuring an automatic compensation for slow shape trends. Interference patterns caused by simultaneously firing units are automatically stored for special analysis. This is a semi-automatic procedure carried out immediately after the fully automatic phase and ensures the correct sorting of interference potentials. When all units have been accounted for, the following histograms can be displayed for each unit type: a simple interval versus time plot; an interval occurrence probability histogram; a joint interval histogram.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call