Fig. 3: Proposed DPN for design parameters from the given absorption spectrum of the meta-absorber. | Light: Science & Applications

Fig. 3: Proposed DPN for design parameters from the given absorption spectrum of the meta-absorber.

From: Machine learning assisted plasmonic metascreen for enhanced broadband absorption in ultra-thin silicon films

Fig. 3

a Schematic of the DPN. b The decrease of training and validation loss during DPN training. c The regression plots of normalized design space where the solid black line represents the best fit and green dots show the predicted parameters. The regression coefficients verify the effectiveness of the trained network as reported in (i)–(vi)

Back to article page