Blind Federated Learning without initial model

Journal of Big Data

Table 4 Blind FL results (constant weights, slope 5, sigmoid)—Breast Cancer dataset

Agent	Size	% 1s	Accuracy	Accuracy	Precision	Precision
Agent	Size	% 1s	Pre-FL	Post-FL	Pre-FL	Post-FL
1	20%	14%	0.6818	0.9091	0.3000	0.6000
2	20%	26%	0.8696	0.8696	0.7143	0.6250
3	20%	35%	0.8696	0.8696	0.7778	0.7778
4	20%	13%	0.9130	0.9565	0.6667	0.5000
5	20%	26%	0.8696	0.9130	0.6667	0.7500
Avg.	–	–	0.8407	0.9036	0.6251	0.6506
1	19%	24%	0.9063	0.9063	0.6667	0.6667
2	26%	35%	0.9048	0.9365	0.9130	0.8846
3	16%	42%	0.7317	0.7317	1.0000	0.9167
4	15%	20%	0.9231	0.9231	0.7500	0.7500
5	24%	47%	0.8889	0.8889	0.9091	0.9091
Avg.	–	–	0.8709	0.8773	0.8478	0.8254
1	5%	27%	0.9167	0.9167	0.8333	0.8333
2	4%	43%	0.8667	0.8667	1.0000	1.0000
3	42%	35%	0.9300	0.9300	0.9189	0.8919
4	20%	23%	0.9487	0.9487	0.8571	0.8571
5	29%	49%	0.8571	0.8571	0.9474	0.9474
Avg.	–	–	0.9038	0.9038	0.9114	0.9059
1	3%	20%	0.8182	0.8182	0.3333	0.2500
2	4%	36%	0.7273	0.8182	0.5714	0.5714
3	45%	40%	0.8660	0.8763	0.8824	0.9032
4	21%	17%	0.9000	0.9250	0.7000	0.7778
5	26%	49%	0.9155	0.9155	1.0000	1.0000
Avg.	–	–	0.8454	0.8706	0.6974	0.7005