Novel neurosymbolic artificial intelligence (NSAI) based algorithm to predict specific energy absorption in CoCrMo based architected materials

In this paper, two neurosymbolic-based methods—the neurosymbolic decision tree (DT) and the neurosymbolic XGBoost—are compared with a simple artificial neural network (ANN) to see how well they perform. This provides a novel comparison of prediction algorithms. The outcomes demonstrate that the neurosymbolic-based algorithms outperform other algorithms in terms of mean squared error (MSE) and R-squared (R²) value. Simple ANN gave values of 8.4 and 0.92, Neurosymbolic Decision tree gave values of 1.4 and 0.98, Neurosymbolic XGBoost gave values of 0.62 and 0.99 as MSE and R² respectively. When combined, symbolic and neurological components offer a new methodology that is more accurate and comprehensible. This work highlights the ways in which neurosymbolic techniques can be applied to improve predictive modeling in several domains, contributing to the growing body of research on the subject.