Figure 3：Under linear TPPP operation, spectrum classification task is performed. NIR spectra of flour samples predicted moisture with Richardson’s iteration for ridge regression on TPPP. (IMAGE)

Light Publishing Center, Changchun Institute of Optics, Fine Mechanics And Physics, CAS

Figure 3：Under linear TPPP operation, spectrum classification task is performed. NIR spectra of flour samples predicted moisture with Richardson’s iteration for ridge regression on TPPP.

Caption

Figure 3：Under linear TPPP operation, spectrum classification task is performed. NIR spectra of flour samples predicted moisture with Richardson’s iteration for ridge regression on TPPP. (a) a Schematic overview of Richardson iteration implemented on TPPP hardware for FT-NIR flour moisture prediction. FT-NIR spectra from flour samples with different moisture levels are compressed by PCA and used for ridge-regression-based prediction. The matrix inversion required by ridge regression is reformulated as a Richardson iteration and implemented on TPPP hardware using the linear-mode reconfigurable kernels. (f) FT-NIR transmission spectra for flours with different moisture content, with corresponding predicted and ground-truth moisture values. (g) Comparison of predicted and measured flour moisture for a static model (blue) and with dynamic adaptation (red). (h) Statistical prediction error metrics, including mean error (μ), standard deviation (σ), root mean squared error (RMSE), and coefficient of determination (R2)

Credit

Qixiang Cheng et al.

Usage Restrictions

Credit must be given to the creator.

License

CC BY

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.