Itogovaya-attestacionnaya-rabota

1.

FB7B64O 4FF5EF4F8BAA4O
D45BF4
=4?<F<> 44==OE: Python < SQL 6 BigData
O?>?=<?4: G74A>B6 8D8?? 4>E8@B68G
@G??4: $ = -802
4@<4=F: 4

2.

=4?8F8G5E>4O 7444G4
$47D45BF4FL 4?7BD8F@ 4?O BF5A>8 E>?BAABEF8 >?85AF4 >
CB?B68F5?LAB@G BF>?8>G A4 D5>?4@AB5 CD54?B65A85 CB 57B
CD87A4>B6B@G BC8E4A8N.
B@C4A8O E@B65F B5D4M4FLEO FB?L>B > F5@ >?85AF4@, >BFBDO5
E>?BAAO > BF>?8>G, BCF8@878DGO @4D>5F8A7B6O5 74FD4FO.

3.

?4A 6OCB?A5A8O CDB5>F4
01
02
47DG7>4 44AAOE
5D68GA4O B5D45BF>4
'F5A85 44AAOE 87 D49?B6
BDD5>F8DB6>4 747B?B6>B6, F8CB6, CDB65D>4 4G5?8>4FB6 8
4AB@4?89
03
04
47654BGAO9 4A4?87
EE?54B64F5?LE>89 4A4?87
OCB?A5A85 EDA 4?O CBA8@4A8O 44AAOE
CD545?5A85 646AOE CD87A4>B6 8 6748@BE6O759
05
06
BEFDB5A85 @B45?8
FB7B6O9 4?7BD8F@
%B744A85 CDB7AB7AB9 @B45?8
$47D45BF>4 E8EF5@O BF5A>8 E>?BAABEF8

4.

5@68G=4O >5@45>F>4
44==OE
@>?GM5==O5
7=4G5=8O
PREVIOUS_CARD_NUM_UTILIZ
ED 74CB?A5A AG?O@8
@>65@>4 4G5?8>4F>6
OCB?A5A4 CDB65D>4 A4
A4?8G85 4G5?8DGNM8EEO
74C8E59 6 44F4E5F5
WORK_TIME 74CB?A5A
@5484AB9
?O C5AE8BA5DB6 WORK_TIME
=0
>@@5>F8@>6>4 F8?>6
4F57BD84?LAO5 C5D5@5AAO5 CD5B5D47B64AO 6 F8C 'category'

5.

$47654BGAO9 4A4?87 44AAOE (EDA)
A4?87 F5?56B9 C5D5@5AAB9
4F57BD84?LAO5 CD87A4>8
7GG5AB D4ECD545?5A85 F5?56B9 C5D5@5AAB9 TARGET 8
EBBFABH5A85 >?4EEB6 4?O CBA8@4A8O 54?4AE4 44AAOE.
DB4A4?878DB64AO C5D6O5 5 >4F57BD84?LAOE CD87A4>B6
E 687G4?874F859 8E D4ECD545?5A8O.
'8E?B6O5 CD87A4>8
BEFDB5AO 78EFB7D4@@O 4?O 6E5E G8E?B6OE CD87A4>B6
4?O 6OO6?5A8O D4ECD545?5A89.

6.

46AO5 CD87A4>8 @B45?8
8GAO9 4BEB4
B?8G5EF6B 45F59
B7D4EF
PERSONAL_INCOME 4
>?NG56B9 D4>FBD E>?BAABEF8
> BF>?8>G
CHILD_TOTAL 6?8O5F A4
D5H5A85 >?85AF4
AGE 4 7A4G8@O9
45@B7D4D8G5E>89 CB>474F5?L
%G@@4 >D548F4
EFBD8O CDBEDBG5>
CREDIT BCD545?O5F
D8A4AEB6O5 6B7@B6ABEF8
LOAN_DLQ_NUM,
LOAN_MAX_DLQ_AMT
CB>47O64NF A4456ABEFL
ECB?L7B64A @5FB4 SelectKBest 4?O BF5BD4 15 A485B?55 7A4G8@OE CD87A4>B6. BEFDB5A4 @4FD8F4 >BDD5?OF89 4?O
4A4?874 6748@BE6O759.

7.

57G?LF4FO @B45?8
0.75-& 100
15
ROC-AUC Score
5D56L56
D87A4>B6
OEB>4O
CD54E>474F5?LA4O
ECBEB5ABEFL @B45?8
Random Forest
B?8G5EF6B 45D56L56
6 4AE4@5?5
FB5D4AB 4?O
D8A4?LAB9 @B45?8
B45?L B5GG5A4 E 8ECB?L7B64A85@ RandomForestClassifier E
54?4AE8DB6>B9 >?4EEB6. OCB?A5AB @4EHF458DB64A85 CD87A4>B6 8
D4745?5A85 44AAOE A4 B5GG4NIGN 8 F5EFB6GN 6O5BD>8.

8.

?NG56O5 6748@>A6O78
OA>>89 4>E>4
>?>4>9 6>7@4AB
%>@>H4O 8AB>@8O
?85AFO E 5B?55 6OEB>8@ 4BEB4B@
G4I5 BF>?8>4NFEO A4 CD54?B65A8O
?85AFO 25-35 ?5F 5B?55 4>F86AO 6
BF>?8>4E
B?B68F5?LA4O >D548FA4O 8EFBD8O
CB6OH45F 65DBOFABEFL EB7?4E8O
@>65@>4 78?>B57O: %F4F8EF8G5E>89 F5EF 4AA4-'8FA8 CB4F65D48?, GFB ?8GAO9 4BEB4 7A4G8@B 6?8O5F A4 BF>?8>
>?85AF4 (p-value < 0.05)

9.

5>B@5A44F88 4?O 587A5E4
!57@5AF4F8O
>?85AFB6
5DEBA4?874F8O
CD54?B65A89
%DB>GE8DB64FLEO A4
>?85AF4E E 65DBOFABEFLN
BF>?8>4 6OH5 0.7 4?O
@4>E8@4?LAB9
MDD5>F86ABEF8 >4@C4A89
'G8FO64FL 4BEB4, 6B7D4EF 8
>D548FAGN 8EFBD8N CD8
DBD@8DB64A88
8A48684G4?LAOE
CD54?B65A89
CF8@874F8O 74FD4F
%A878FL @4D>5F8A7B6O5
D4EEB4O A4 >?85AFB6 E
A87>B9 65DBOFABEFLN
BF>?8>4
%B744A >?4EE CustomerResponsePredictor 4?O CD4>F8G5E>B7B
CD8@5A5A8O @B45?8 6 587A5E-CDBF5EE4E >B@C4A88.