. 2022 Jun 2;3(6):100521. doi: 10.1016/j.patter.2022.100521

Table 2.

Performance for federated molecular regression

Dataset	α	Centralized training		Federated learning
Dataset	α	MolNet^a	FedChem^a_ours	FedAvg	FedProx	MOON	FedFocal_ours	FedVAT_ours	FLIT_ours	FLIT+_ours
FreeSolv $⇓$	0.1	1.40	1.430	1.771	1.693	1.376	1.686	1.371	1.634	1.228^d
	0.5			1.445	1.376	1.423	1.322	1.299	1.366	1.127^d
	1			1.223	1.216	1.469	1.294	1.150	1.277	1.061^d
Lipophilicity $⇓$	0.1	0.655	0.6290	0.6361^d	0.6403	0.6426	0.6403	0.6556	0.6563	0.6392
	0.5			0.6306	0.6365	0.6339	0.6351	0.6333	0.6368	0.6270^d
	1			0.6505	0.6474	0.6442	0.6461	0.6488	0.6443	0.6403^d
ESOL $⇓$	0.1	0.97	0.6570	0.8016	0.7702	0.7537^d	0.8022	0.7776	0.7788	0.7642
	0.5			0.7524	0.7382	0.7258	0.7708	0.7243	0.7426	0.7119^d
	1			0.7056	0.6828	0.6751	0.6822	0.7253	0.6705^d	0.6998
QM9 $⇓$	0.1	0.0479^b	0.0890^c	0.5889	0.6036	0.5817	0.6164	0.5606	0.5713	0.5356^d
	0.5			0.5906	0.5751	0.5707	0.6059	0.5656	0.5658	0.5222^d
	1			0.5786	0.5691	0.5808	0.5822	0.5602	0.5621	0.5282^d

$⇓, ⇑$ indicate if lower or higher numbers are better.

Results were obtained with centralized training.

Results were retrieved from Klicpera et al.² with a seperate SchNet for each task.

Results were obtained by a single multitask network. Smaller α of LDA generates more extreme heterogeneous scenario. FedFocal and FedVAT are proposed in this paper as the variants of FLIT(+).

Best federated-learning results.