-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathREADME.md_Bank Credit-Score.rtf
81 lines (81 loc) · 3.56 KB
/
README.md_Bank Credit-Score.rtf
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
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
{\rtf1\ansi\ansicpg1252\deff0\nouicompat\deflang1033{\fonttbl{\f0\fnil\fcharset0 Arial Black;}{\f1\fnil\fcharset0 Times New Roman;}{\f2\fnil\fcharset1 Cambria Math;}{\f3\fnil Times New Roman;}{\f4\fnil Arial Black;}}
{\colortbl ;\red0\green0\blue255;}
{\*\generator Riched20 10.0.17134}{\*\mmathPr\mmathFont2\mwrapIndent1440 }\viewkind4\uc1
\pard\sa200\sl276\slmult1\b\f0\fs40\lang9 Data Science PROJECT \par
\fs28 Client: Bank GoodCredit | Category: Banking - Risk \fs40\par
\b0\f1\fs28 This project contains only brief information of the project for more information E-mail me @ {{\field{\*\fldinst{HYPERLINK http://[email protected] }}{\fldrslt{http://[email protected]\ul0\cf0}}}}\b\f0\fs28\par
\ul\f1 Business Case: \b0\par
\ulnone Bank wants to predict credit score for current credit card customers. The credit score will denote a customer\rquote s credit worthiness and help the bank in reducing credit default risk. \par
\ul\b Target variable \ulnone\b0\f2\u8594?\f1 \par
Bad_label:\par
0 \f3\endash\f1 Customer has Good credit history\par
1 \f3\endash\f1 Customer has Bad credit history \b\f4 \par
\ul\f1\lang16393 ASSUMPTIONS:\par
\ulnone\b0 Provided Data set is imbalanced\par
\ul\b Existing Data set:\par
\ulnone\b0 Client provided Customer accnt , Enquiery Details , Customer_Demographics\par
\tab -Bussiness requirement wants to predict the Target variable wheather the customer has Good credit history or Bad credit history \par
\ul\b Steps to predict the Good or Bad credit history:\par
\ulnone\b0 Data set is provided by SQL server with USERNAME<PASSWORD<HOST<PORT by the clint.\par
\ul Step 1:\par
\ulnone\tab import the data set to jupiter and convert it to CSV file.so that it will become convinent to analyze the data.\par
\ul Step 2: \ulnone\par
\tab After converting to csv , open the fie in xl-sheet or in Tableau\par
\ul Step 3\ulnone :\par
\tab Analyze the data clenly and amke out your seperate table to take into consideration\par
\ul DATA CLEANING & DATA MINING (in brief)\par
\ulnone 1.Read the CSV_file\par
2.Remove the irrelevant columns\par
3.Replace the data with ['?','*','$',' ',' ',''] with Nan\par
4.Drop the duplicate columns\par
Enquiery File Transform(Data Mining)\par
1.Read the CSV of cust_enquiery\par
2.Take relevant columns \par
3.Add one more column of enq_eqn_amt(which add the no of transction made by every single customer)\par
4.Groupby customer_no(which is included in all 3 dataset)\par
5.join the customer_no and df_amt,df_count\par
6.save the file\par
Account File Transformation\par
1.Read the file\par
2.Take the relevant columns\par
3.fill NaN with '0'\par
4.Group by Customer_no\par
5.Save the file\par
>\ul JOINING THE DATA\par
\ulnone 1.join all the 3 files with left join\par
\ul Processing the Data\par
\ulnone 1.Convet the categorical data to numerical data by label Encoder \par
2.Take Bad_label as Target Variable/Dependent Variable\par
3.Other columns are independent Variable\par
4.Train & test the Data from sklearn.model_selection\par
Applying ML algorithm\par
\tab Fit the model with Random Forest algorithm and after traing with Random Forest, find out accuracy and y_predict\par
Plot the Graph\par
\tab plot the graph with feature importance and with independent Variables\par
\ul Result\ulnone :\par
Achived 96.13% accuracy by Random Forest model fit method\par
\ul Tools Used:\ulnone\par
1.Sql server\par
2.Excel\par
3.Jupyter\par
4.ML Algorithm\par
\par
\par
\par
\par
\par
\par
\par
\par
\par
\par
\par
\par
\par
\par
\par
\par
\par
\b\f0\fs40\lang9\par
}