ANN in Science: Entomology., pp. 23-32. In, Havel J. & Vaňhara J., Redes neuronales en Ciencia.
|Year of publication||2009|
|MU Faculty or unit|
|Description||Artificial Neural Networks (ANN) belonging to the Artificial Intelligence methods play an ever increasing role in modern science. Developed in the 1950s, inspired by the neuron structure and the way how human brain works they have been finding increasingly powerful and exciting applications in all branches of science. While via hard model chemical or biological systems can be exactly described by formulas, equations and the values of parameters, ANN using a soft model can do the same even when the exact description is not known or is too complex. ANN are often thought to be something mysterious, very difficult to understand and therefore presented just as a black box. Therefore the principals of ANN will be explained and their enormous potential for modelling of a broad range of processes fitting under virtually all areas of science will be elucidated. Utilisation of ANN will be documented on abundant examples from numerous areas of science and chemistry including analytical chemistry which is the speaker s primary field of expertise. The examples involve applications in classification, biology, medical diagnosis, forensic science, soft modelling of various chemical processes, QSAR, optimisation, multicomponent analysis, process analysis, and optimization of analytical methods for the determination of a broad range of analytes including simple ions, antiviral drugs, pharmaceutical products, antidotes against chemical weapons, nucleotides, complex peptides/protein mixtures etc.|