Skip to main content
. 2020 Mar 25;20(7):1813. doi: 10.3390/s20071813

Figure 8.

Figure 8

A comparison between regression predictions during honey-water blending for the central sensor, non-central sensor, and combining the outputs from both sensors. A time domain input CNN was used as the learning algorithm. R2 values: 0.825 non-central sensor, 0.936 central sensor, and 0.975 multisensor data fusion.