Abstract

Artificial neural networks (ANNs) are the result of academic investigations that use mathematical formulations to model nervous system operations. Neural networks (NNs) represent a meaningfully different approach to using computers in the workplace, and have been used to recognize patterns and relationships in data. In this paper, the compressive strength (CS) of lightweight material with 0, 20, 30, and 50% of scoria instead of sand, and different water-cement ratios and cement content for 288 cylindrical samples were studied. Out of these, 36 samples were randomly selected for use in this research. The CS of these samples was used to teach ANNs CS prediction to achieve the optimal value. The ANNs were formed by MATLAB software so that the minimum error in information training and maximum correlation coefficient in data were the ultimate goals. For this purpose, feed-forward back propagation (FFBP) with TRAINBR training function, LEARNGD adaption learning function, and SSE performance function were the last networks tried. The end result of the FFBP was 3-10-1 (3 inputs, 10 neurons in the hidden layer, and 1 output) with the minimum error below 1% and maximum correlation coefficient close to 1.    

Key words: Compressive strength (CS), scoria, artificial neural networks (ANNs), feed-forward back propagation (FFBP).