Fish-school species identification using a neural network <B>Fish-school species identification using a neural network</B><P> Haralabous J. and Georgakarakos S.</B><I><P> Institute of Marine Biology of Crete (IMBC), Department of Hydroacoustics, P.O. Box 2214, GR-710 01 Heraklion, Crete, Greece</I><B><P> Summary</B><I></I><P> Fish-schools of Sardines, Anchovies and Horse-Mackerel can be discriminated from each other, using processed data from hydroacoustic surveys. Back-propagation artificial neural networks can be trained to reliably classify such schools, even in the presence of significant overlaps in the species' school characteristics.<B><P> Reference:</B> International Council for the Exploration of the Sea. Fish Capture Committee. CM 1993/ B:9. Session T<P>