Fig. 1. Machine learning methods generally used in additive manufacturing

Fig. 2. Melt pool temperature distribution predicted by PINN^[23]. (a) A framework for the prediction of melt pool temperature and dynamics by PINN; (b) comparison of temperature and melt pool prediction results by finite element method, PINN, and experiment

Fig. 3. Monitoring the formation of surface defects during DED by in situ cloud processing and machine learning method^[29]

Fig. 4. Prediction of processing window of additive manufacturing by supervised machine learning^[34]

Fig. 5. Machine learning prediction of porosity in additively manufactured 17-4 PH steel^[13]. (a) Prediction of porosity versus laser machining parameters; (b) standard error of predicted value

Fig. 6. Visualization of Gaussian process regression model^[37]. (a) Relative density of SLM-processed AlSi10Mg with different processing parameters; (b) relative density value predicted by GPR model; (c) density error value predicted by GPR model

Fig. 7. Optimization processes of DED deposition toolpath assisted by machine learning^[21]

Fig. 8. Machine learning assisted design of high strength and tough aluminum alloys^[49]. (a) Relationship between elongation and ultimate tensile strength; (b) relationship between fracture toughness and ultimate tensile strength

Fig. 9. Schematic of mechanistic data-driven framework for mechanical property prediction^[60]

Fig. 10. Machine-learning-assisted composition design, process optimization, and performance control in additive manufacturing metallic materials^{[37,44,58,61]}