Identifying Potential Clinical Syndromes of Hepatocellular Carcinoma Using PSO-Based Hierarchical Feature Selection Algorithm

Table 1

The result of feature ranking for all the syndromes.

Syndrome (𝑓𝑖)	Name of syndrome	Abbreviation	Size	mcc(𝑓𝑖,𝑝𝑠)	pe(𝑓𝑖)	fr(𝑓𝑖)	Rank
1	Lip color	LC	4	0.9257	0.1688	0.9480	24
2	Tongue color	Tc	4	0.9293	0.1813	0.9487	25
3	Appearance of tongue-1	At1	3	0.8123	0.5808	0.8550	16
4	Appearance of tongue-2	At2	5	0.9712	0.2998	0.9592	27
5	Coated tongue color	Ctc	3	0.8589	0.1914	0.9126	21
6	Texture of coated tongue	Tct	7	0.9039	0.2518	0.9298	23
7	Position of coated tongue	Pct	5	0.9629	0.2685	0.9578	26
8	The color of complexion	Coc	8	0.6396	2.7350	0.5790	6
9	Whole body condition	Wbc	8	0.9326	1.0378	0.8749	19
10	Odor	Od	1	0.6948	0.6055	0.7941	13
11	Chilly	Ch	1	0.6011	0.4767	0.7586	10
12	Hectic fever	Hf	1	0.7890	0.4248	0.8571	17
13	Fever	Fe	1	0.7304	0.2969	0.8391	15
14	Sweating	St	2	0.6270	0.4875	0.7706	11
15	Facial features	Ff	13	0.2177	5.6792	0.1088	1
16	Cardiothoracic condition	Ca	4	0.4923	1.2036	0.6402	8
17	Sternocostal and abdominal pain	Sap	16	0.4010	1.6943	0.5513	5
18	Diet	Diet	7	0.2937	1.7266	0.4948	3
19	Defecate and urine	Du	10	0.4016	1.7268	0.5488	4
20	Sleep	Slp	2	0.4382	0.9854	0.6324	7
21	Emotion	NEs	3	0.5141	1.1600	0.6549	9
22	Skin of the limbs	Sl	10	0.2091	2.4312	0.3905	2
23	Bump in ribs	Bir	1	0.6543	0.5541	0.7784	12
24	Ascites	Ass	1	0.7279	0.4304	0.8260	14
25	Pleural effusion	Pe	1	0.7894	0.2301	0.8745	18
26	Pulse condition in left	Pcle	13	0.8630	0.3003	0.9051	20
27	Pulse condition in right	Pcrt	13	0.8716	0.2556	0.9133	22