High strength aluminum alloys design via explainable artificial intelligence

Accelerated discovery of high-performance Al-Si-Mg-Sc casting alloys by integrating active learning with high-throughput CALPHAD calculations

“…Machine learning modeling aims to establish a function between input and output and makes it as close to the real function relationship as possible by optimizing the model parameters [ 50 , 51 ]. Due to its low computational cost and short development cycle, machine learning is coupled with powerful data processing and high prediction performance and is being widely used in material science, including the establishment of phase diagrams [ 52 ], properties prediction [ 53 ], the discovery and design of high-performance materials [ 38 , 54 ], and the exploration of strengthening and toughening mechanism [ 38 , 40 ]. The widely used machine learning algorithms include linear algorithms, decision tree-based (DT) algorithms, artificial neural network (ANN), support vector machines (SVM), random forest (RF), and some Bayesian-based algorithms [ 55 ].…”

“…With the development of computer technology, various computational assisted alloy design methods, like first-principles (FP) calculations [ 33 ], molecular dynamics (MD) simulations [ 34 ], computational thermodynamics (CT) [ 11 , 15 ], computational kinetics [ 35 , 36 ], phase-field (PF) simulations [ 37 ], and machine learning (ML) approach [ 38 , 39 ], have been widely used to accelerate the development of high-performance alloys. The CT method, which can construct the quantitative relationship between the composition and microstructure of alloys, has been recently used to efficiently design the optimal Sc contents in i.e.…”

Review on the Use of Machine Learning Techniques to Optimize the Processing of Copper Alloys in Additive Manufacturing

“…Experience from successful experiments 3,4 has shown that the proven approach is to first determine the original hyperparameters through a screening process and subsequently use Bayesian optimization to solve the problem. The reason why Bayesian optimization is commonly used is that by using this method, the optimal values of the independent variables and their corresponding dependent variables can be eventually obtained by guessing, without being sure of the exact form of the functionand faster than other methods.…”