Table 3.

Performance of Naïve Model and the snapshot ensemble after combining various snapshots on the test database.

Combinations	Accuracy (%)
Naïve model	82.836
S1	86.16
S2	72.39
S3	87.69
S4	87.69
S5	85.08
(S1, S2)	84.328
(S1, S3)	87.313
(S1, S4)	87.313
(S1, S5)	88.060
(S2, S3)	87.313
(S2, S4)	86.567
(S2, S5)	85.075
(S3, S4)	86.940
(S3, S5)	88.060
(S4, S5)	87.687
(S1, S2, S3)	87.313
(S1, S2, S4)	87.687
(S1, S2, S5)	86.194
(S1, S3, S4)	88.060
(S1, S3, S5)	88.060
(S1, S4, S5)	87.687
(S2, S3, S4)	88.060
(S2, S3, S5)	86.567
(S2, S4, S5)	87.313
(S3, S4, S5)	88.060
(S1, S2, S3, S4)	87.687
(S1, S2, S3, S5)	88.433
(S1, S2, S4, S5)	88.060
(S1, S3, S4, S5)	87.687
(S2, S3, S4, S5)	87.687
(S1, S2, S3, S4, S5)	87.313

Bold value indicates highest result among all the snapshot conditions.