Correlation-Based Ensemble Feature Selection Using Bioinspired Algorithms and Classification Using Backpropagation Neural Network

Table 1

Outline of hepatitis datasets.

S. no.	Feature	Description	Datatype
1.	Age	Age of the patient	Numerical
2	Sex	Gender of the patient	Categorical
3	Steroid	Whether the patient has taken anabolic steroids or not	Boolean
4	Antivirals	Whether the patient has taken antivirals or not	Boolean
5	Fatigue	Whether the patient has experienced extreme tiredness or not	Boolean
6	Malaise	Whether the patient is having a vague feeling of body discomfort	Boolean
7	Anorexia	Whether the patient has lack or loss of appetite for food	Boolean
8	Liver big	Whether the patient’s liver is enlarged or not	Boolean
9	Liver firm	Whether the patient’s liver is firm or not	Boolean
10	Spleen palpable	Whether the patient’s spleen is enlarged or not	Boolean
11	Spiders	Whether the blood vessels are near the skin surface due to the increased estrogen level.	Boolean
12	Ascites	Whether the fluid is accumulated in the peritoneal cavity or not	Boolean
13	Varices	Whether the patient is having bleeding from varices or not	Boolean
14	Bilirubin	The amount of bilirubin in the blood sample	Numerical
15	Alk phosphate	Level of alkane phosphate in the blood sample	Numerical
16	Sgot	The amount of serum lutamic oxalo acetic transaminase in the blood	Numerical
17	Albumin	The amount of serum albumin protein in the clear liquid portion of the blood sample	Numerical
18	Protime	Time taken for blood plasma to clot	Numerical
19	Histology	Class attribute indicates whether the patient survives or not	Boolean