CS-1 : Features Descriptions, Engineering & Results

	A	B	C	D	E	F
1	Feature Name	Feature Description	Present in Actual dataset?	Logic	Reason behind adding new feature	Note
2	Bene ID	It contains the unique id of the beneficiary.	Y	N/A	N/A	N/A
3	DOB	It contains the date of birth of the beneficiary.	Y	N/A	N/A	N/A
4	DOD	It contains the date of death of the beneficiary.	Y	N/A	N/A	N/A
5	Gender	It represents the gender of the beneficiary.	Y	N/A	N/A	N/A
6	Race	It represents the human race of the beneficiary.	Y	N/A	N/A	N/A
7	State	It represents state code in which a beneficiary lives or belongs to.	Y	N/A	N/A	N/A
8	Country	It represents country code in which a beneficiary lives or belongs to.	Y	N/A	N/A	N/A
9	Renal Disease Indicator	It represents a code which indicates whether a beneficiary had a long history of kidney disease at the time of buying a plan.	Y	N/A	N/A	N/A
10	Chronic Condition Indicators * 1,2,3,4,5,6,7,8,9,10,11	It represents a code which indicates whether a beneficiary had a chronic condition of a specific disease at the time of buying a plan.	Y	N/A	N/A	N/A
11	NoOfMonths_PartACov	It represents number of months of part-A coverage.	Y	N/A	N/A	Dropped Because it doesn't have any variability.
12	NoOfMonths_PartBCov	It represents number of months of part-B coverage.	Y	N/A	N/A	Dropped Because it doesn't have any variability.
13	IPAnnualReimbursementAmt	It consists of the maximum reimbursement amount allocated to a beneficiary for annual hospitalization on the basis of insurance plan.	Y	N/A	N/A	N/A
14	IPAnnualDeductibleAmt	It consists of the maximum co-payment to be bourne by a beneficiary annualy if in case gets hospitalized.	Y	N/A	N/A	N/A
15	OPAnnualReimbursementAmt	It consists of the maximum reimbursement amount allocated to a beneficiary for annual non-hospitalization on the basis of insurance plan.	Y	N/A	N/A	N/A
16	OPAnnualDeductibleAmt	It consists of the maximum co-payment to be bourne by a beneficiary annualy if in case only visited a hospital without admission.	Y	N/A	N/A	N/A
17	Bene_Age	Age of a beneficiary.	N	(DOD - DOB) / 365	To check whether BENE Age has some influence on the fraudulent cases?	Added
18	Age_Groups	Age group to which a beneficiary belongs.	N	-	To check whether a particular BENE Age Group has some influence on the fraudulent cases?	Dropped Initially, I added this feature but, later on found it to be providing similar information like PRV_Bene_Age_Sum. Thus, dropped it.
19	PRV_Tot_Unq_DOB_Years	Year of birth of a beneficiary.	N	Year of DOB	The idea behind adding this feature is that if a provider has very high variability in the year of birth of patients then that might be one of the signs of medicare frauds.	Added What didn't worked? The raw DOB Year feature doesn't have much variability, thus added the above mentioned feature.
20	N/A	Month of birth of a beneficiary.	N	Month of DOB	To check whether BENE Month of birth has some influence on the fraudulent cases?	Not added What didn't worked? This feature doesn't have any variation.
21	N/A	Year of death of a beneficiary.	N	Year of DOD	To check whether BENE Year of death has some influence on the fraudulent cases?	Not added What didn't worked? This feature doesn't have any variation.
22	N/A	Month of death of a beneficiary.	N	Month of DOD	To check whether BENE Month of death has some influence on the fraudulent cases?	Not added What didn't worked? This feature doesn't have any variation.
23	Is_Alive?	Whether a beneficiary is alive or not?	N	0 if DOD is not null else 1	To check whether BENE Life Status has some influence on the fraudulent cases?	Added
24	PRV_Bene_Age_Sum	Sum of patients age treated by a Provider	N	-	The idea behind adding this feature is that there might be a pattern like if the sum of patients age treated by a provider is very high or low then it might influence the fraud.	Added
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100