Fig. 4: Sensitivity analysis of E2T for the two extrapolative prediction tasks (HOIP-GeF and HOIP-PbI) using the HOIP dataset. | Communications Materials

Fig. 4: Sensitivity analysis of E2T for the two extrapolative prediction tasks (HOIP-GeF and HOIP-PbI) using the HOIP dataset.

From: Advancing extrapolative predictions of material properties through learning to learn using extrapolative episodic training

Fig. 4

a, b RMSE variations according to varying training support sizes with the inference support size fixed at 1248 (a) and 1228 (b). c, d RMSE variations according to varying the inference support size at λ = 100. In (a, b), the colored lines indicate different smoothing parameters λ, whereas in (c, d), they represent different training support sizes. The shaded areas indicate the standard deviations.

Back to article page