Bu çaly?mada; 4 farkly oranda cam lif ilave edilerek elde edilen 300 ve 350 dozlu betonlara filler malzeme ile hacimce %25, 50, 75 ve 100 oranlarynda yer de?i?tirecek ?ekilde atyk mermer tozu ilave edildi. Elde edilen numunelerin ultrases geçi? hyzlary, porozite de?erleri, basynç dayanymlary ve yarmada çekme dayanymlary belirlendi. Yapay sinir a?yna (YSA) girdi seti olarak dozaj, agrega miktary, lif orany, mermer tozu orany, porozite, ultrases geçi? hyzy ve basynç dayanymy de?erleri seçilerek yarmada çekme dayanymlary tahmin edildi. Çaly?ma sonucunda, geli?tirilen YSA modeli ile deneysel olarak elde edilen veriler kar?yla?tyryldy ve sonuçlaryn uyum içerisinde oldu?u belirlendi.

In this study, filler material and %25, 50, 75 and 100 volume ratio waste marble dust which can be used in place of filler material, added to 300 and 350 dosages concrete obtained by adding glass fibre in four diffferent ratios. Splitting tensile strength, compressive strength, porosity and ultrasonic pulse velocity of obtained sample is detected. Approximate estimation of splitting tensile strength performed via selecting dosages, amount of aggregate, fibre ratio, marble dust ratio, porosity, ultrasonic pulse velocity and splitting tensile strength as an input of Artificial Neural Network (ANN). As a result of this study, developed ANN model compared with the obtained experimental values and it was detected that results match with each other.