Discussion on “Prediction of ground vibrations resulting from the blasting operations in an open-pit mine by adaptive neuro-fuzzy inference system”

Abstract

The discusser appreciates the exhaustive study undertakenby the authors for predicting ground vibrations resultingfrom the blasting operations in an open-pit mine by adap-tive neuro-fuzzy inference system. However, the presentpaper has some scientific mistakes especially in the basicconception and application of the fuzzy theory. Thestraightforward application of ANFIS through softwaregives rise to physical and logical problems even though theresults may seem convenient from the numerical tests pointof view. In any scientific research the first steps shouldinclude the logical and physical principles of the data athand in addition to its simple treatment through somegraphs such as the scatter diagrams prior to classicalregression or ANFIS methodology application. Otherwise,complete dependence on any methodology especially bysoftware without knowing its internal domain may lead toerroneous scientific applications. The paper by the authorssuffers from such defects even though it seems numericallysound. The following logical points may be raised on thescientific grounds.1. Membership functions (MFs) both in Figs. 6 and 7 aremisleading and they are not physically and logicallyplausible. For instance in Fig. 7, ‘‘VH’’ and ‘‘H’’ MFsare almost equivalent within say,