BMC Bioinformatics

Table 4 Model performance comparison

From: A deep learning approach for orphan gene identification in moso bamboo (Phyllostachys edulis) based on the CNN + Transformer model

Model	BA	GM	BM	MCC	Train time (min)	Test time (s)
Random forest	0.667	0.629	0.334	0.227	4	22
SVM	0.690	0.659	0.380	0.252	23	77
RNN	0.517	0.512	0.034	0.245	31	123
CNN + RNN	0.503	0.500	0.007	0.109	18	98
LSTM	0.829	0.829	0.659	0.418	36	284
CNN + LSTM	0.775	0.772	0.550	0.376	26	231
GRU	0.838	0.834	0.667	0.423	33	253
CNN + GRU	0.777	0.776	0.554	0.373	24	219
Transformer	0.844	0.838	0.678	0.444	57	387
CNN + Transformer	0.875	0.871	0.746	0.471	44	343

Bold values indicate the highest values of different evaluation indicators

Back to article page

ISSN: 1471-2105

Contact us

General enquiries: journalsubmissions@springernature.com