Fig. 2: Constructing a prediction model via active learning loops, data augmentation, and collaborative robots.
From: Machine intelligence accelerated design of conductive MXene aerogels with programmable properties
a Schematic illustration of a multi-stage AI/ML framework for constructing a prediction model via active learning loops, data augmentation, and robot-human teaming. b An autonomous testing platform integrated with a UR5e robotic arm and an Instron compression tester. 2D Voronoi tessellation diagrams (c) without and (d) with the GA incorporation after 8 active learning loops. e the mean absolute error (MAE),top, and the mean relative error (MRE), bottom, values of various prediction models based on linear regression, decision tree, gradient-boosted decision tree, random forest, and artificial neural network (ANN) algorithms. f MAE (top) and MRE (bottom) values of various ANN models based on different virtual-to-real data ratios.
