Regional Models for Biological Processes Based on Linear Regression and Neural Networks
Pero J. Radonja, Srdjan S. Stanković, Bratislav Matović, D. Dražić · 2006
In this paper linear regression and neural networks are used for obtaining regional models of biological processes. Regional models enable getting the most important regional characteristics without detailed measurements on all individual objects. Testing of the obtained regional model by using data samples is done. A very high correlation is obtained between real data and data computed on the basis of regional models. It is shown that application of NNs provides better regional models than those obtained by linear regression