Reply to the comments by R. N. Celik and H. S. Kutoglu on Kavzoglu and Saka (2005) 'Modelling local GPS/levelling geoid undulations using artificial neural networks' J Geod 78(9):520-527

摘要

We wish to thank R. N. Celik and H. S. Kutoglu for their interest in and their comments on our paper that we submitted to Journal of Geodesy in May 2003 (Kavzoglu and Saka 2005). We are grateful to the Editor-in-Chief for giving us an opportunity to reply to these comments. First of all, Celik and Kutoglu claim that we conducted a one-sided study in favour of artificial neural networks (ANNs). They criticize us with words that we certainly did not use in Kavzoglu and Saka (2005). To be exact, we did not claim that ANN method is a new method, or we produced the best results with this method. We emphasized that the ANN method can be used as an alternative and more research is needed on this subject. The use of ANNs requires expertise and experience since many parameters, known as internal and external parameters, are set by the analyst. We provided all the details of the setup of the ANNs in the paper (Kavzoglu and Saka 2005, p. 525) in terms of the parameter settings and the characteristics of the data sets.