Concrete is the main choice in the construction of structures due to its high compressive strength. An appropriate method is needed in predicting the compressive strength of concrete so that it can facilitate the planning of the concrete mixture before laboratory testing is carried out. The purpose of this study is to create a model to predict compressive strength using the artificial neural network (ANN) method. The da ta used in the form of concrete compressive strength data from laboratory

more »

... sts collected by Yeh (1998). The concrete mixture consists of cement, water, fine aggregate, coarse aggregate which is grouped into 3 models based on the additives used, namely the ANN-A model with blast furnace slag, fly ash, and superplasticizer additives, the ANN-B model with blast furnace slag additives and the ANN-C model with blast furnace slag and superplasticizer additives. The Modeling of ANN is performed by using MatLab R2021a software. It is carried out in two stages, namely training which uses 80% of the data and validating which uses 20% of the data. The results of the modeling show that the prediction accuracy can be seen in the resulting mean absolute percentage error (MAPE).