Download as pptx, pdf, or txt
Download as pptx, pdf, or txt
You are on page 1of 103

CREDIT EDA

Case Study
Omkar Pednekar
Outliers for Numeric Variable
Outliers for Numeric Variable

 ‘CNT_CHILDREN’ are having most values in first quartile.


 ‘AMT_INCOME_TOTAL’ and ‘DAY_EMPLOYED’ are having
single value outlier.
 ‘AMT_CREDIT’ and ‘AMT_ANUALY’ are having most of the
values in outlier.
 ‘DAYS_REGISTATION’ 50% of values in First quartile.
Univariate Analysis for Contract type
Univariate Analysis for Contract type

 Defaulter are having Cash loan are more than Revolving


loan.
 Non-Defaulter are having also cash loan.
Univariate Analysis of Code Gender
Univariate Analysis of Code Gender

 Females are facing Payment difficulties in Defaulter.


 In Defaulter and Non-Defaulter are having more numbers of
Females facing difficulties.
Univariate Analysis for Flag own car
Univariate Analysis for Flag own car

 Less number of client are have car in Defaulter.


 50% of client are having car in Non-Defaulter
Univariate Analysis of own Realty
Univariate Analysis of own Realty

 High amount of client are having Realty.


 Client not having realty are 50% of client having realty.
Univariate Analysis for types of Suite
Univariate Analysis for types of Suite

 ‘Unaccompained’ type of suite client are large amount of


Defaulter.
 ‘Other_A’ and ‘Other_B’ are having less number of client in
Defaulter.
Univariate Analysis for Income Type
Univariate Analysis for Income Type

  In Defaulter, large amount of working client. less number of


client with Maternity leave client.
 In Non-Defaulter, Student ,unemployed client are having
high risks. they have no Income.
Univariate Analysis for Education Type
Univariate Analysis for Education Type

 Client with ‘Secondary/Secondary special’ Education are


large group in Defaulter
 Client with ‘Academic degree’ are less in Defaulter.
Univariate Analysis for Family Status
Univariate Analysis for Family Status

 Client who are Married are higher Percentage of Defaulter.


 ‘widow’ client are less percent of defaulter.
Univariate Analysis for Housing type
Univariate Analysis for Housing type

 Client already having a home or apartment are higher


percentage in Defaulter.
 Client having ‘co-op-apartment’ and ‘office apartment’
having less amount of defaulter.
Univariate Analysis for Occupation type
Univariate Analysis for Occupation type

 client with Laborers occupation are large no of Defaulters.


 Client with IT staff and HR staff are less in defaulter.
Univariate Analysis for Income Range
Univariate Analysis for Income Range

 client Income range between 100000-200000 are high


percentage of Defaulter.
 client Income range above 400000 are less no of Defaulter.
Univariate Analysis for Credit range
Univariate Analysis for Credit range

 client Credit range between 200000-300000 are high


percentage of Defaulter.
 client Credit range below 100000 are less no of Defaulter.
Univariate Analysis for Age group
Univariate Analysis for Age group

 25-30 age group of client are high percentage in Defaulter.


 60-65 age group client are less percentage.
Bivariate Analysis for Credit vs Education
Bivariate Analysis for Credit vs Education

 client with higher education are having high amount of


Credits in Non Defaulter.
 client with lower secondary are have less credit in Non
Defaulter.
Bivariate Analysis for Income vs Education
Bivariate Analysis for Income vs Education

 higher educated client are having high Income in non


Defaulter.
 lower secondory are having low Income in non defaulter.
Bivariate Analysis for Credit vs Family Status
Bivariate Analysis for Credit vs Family Status

 married client are high Credits in Non Defaulter.


 Single/not married are having more no. of outlier.
Bivariate Analysis for Income vs Family Status
Bivariate Analysis for Income vs Family Status

 ‘married’ have more outlier.


 Family status variable most values in 3rd quartile.
Bivariate Analysis for Credit vs Occupation
Bivariate Analysis for Credit vs Occupation

 ‘Manager’ having more outier.


 ‘Accountant’ and ‘Manager’ are high in Credit.
Bivariate Analysis for Income vs Occupation
Bivariate Analysis for Income vs Occupation

 ‘Manager’ have higher Income.


 ‘manager’ have more outlier.
 low-skill laborers are low in income amount.
Bivariate Analysis for Credit vs Gender
Bivariate Analysis for Credit vs Gender

 Females and Males are almost same Credit amount.


 Male and Females Credit value in 1st Quartile.
Bivariate Analysis for Income vs Gender
Bivariate Analysis for Income vs Gender
 Famale and males Income values are in 3rd Quartile.
 Males are more outliers.
Bivariate Analysis for Credit vs Contract Type

 client having Cash loan are high in Credit amounts


Bivariate Analysis for Income vs Contract Type


  cash loan client are high in Income as compare to revolving loans.
Bivariate Analysis for Credit vs Age group

 40-45 and 45-50 age group are having high Credits.


Bivariate Analysis for Income vs Age group

 40-45 age group are high in Income.


60-65 age group are low in Income.
Bivariate Analysis for Credit Amount vs Education

 defaulter client are high in Credit with Secondary/secondary


special
Bivariate Analysis for Income vs Education

 higher education client having high income in Defaulter.


Bivariate Analysis for Credit vs Family Status

 married client are having high credit in Defaulter.


Bivariate Analysis for Income vs Family Status

 married client are high in income in Defaulter.


Bivariate Analysis for Credit vs Occupation

 accountant,manager and laborers are high in Credit.


Bivariate Analysis Income vs Occupation

 managers are having high income in Defaulters


Univariate Analysis for Credit vs Gender

 Females are high credits in Defaulter


Univariate Analysis for Income vs Gender

 Females are having high Income in Defaulters.


Bivariate Analysis for Credit vs Contract Type

 client with cash loans are have high Credits


 Cash loan have more outlier
Bivariate Analysis for Income vs Contract Type

 client with cash loans are having high Income.


 Cash loan are more outlier
Bivariate Analysis for Credit vs Age group

 50-55 age group clients having high credits.


Bivariate Analysis for Income vs Age group

 30-35 age group client having high Income.


Multivariate Analysis for Credit Amount vs Education vs Family Status
Multivariate Analysis for Credit Amount vs
Education vs Family Status
 higher education of family status of 'marriage', 'single' and
'civil marriage' are having more outliers.
 Civil marriage for Academic degree is having most of the
credits in the third quartile
 Academic degree education with family status of 'Civil
marrage‘ ,'separated‘ and 'married' are having high Credit.
Multivariate Analysis for Income Amount vs Education vs
Family Status
Multivariate Analysis for Income Amount vs
Education vs Family Status

 'Higher Education' and 'Secondary/secondary special' are


having more outliers.
 'Academic degree' client are having high Income.
Multivariate Analysis for Credit Amount vs Education vs Family Status
Multivariate Analysis for Credit Amount vs
Education vs Family Status
 'secondary/secondary special' are having more outliers.
 'academic degree' education with married client having high
Credits.
Multivariate Analysis for Income Amount vs
Education vs Family Status
Multivariate Analysis for Income Amount vs
Education vs Family Status
 'secondary/secondary special' education are having
more outliers.
'academic degree' education married client having
high income.
Multivariate Analysis for Credit Amount vs Education vs Family Status
Multivariate Analysis for Credit Amount vs
Education vs Family Status
 'Higher education' have more outliers.
'married‘ , 'civil marriage‘ and 'separated‘
 client with 'academic education' having high Credits
Multivariate Analysis for Income Amount vs Education vs
Family Status
Multivariate Analysis for Income Amount vs
Education vs Family Status
 'higher education' and 'secondary/secondary special' having
more outliers.
 'married', 'single', 'civil marriage' and 'separated' client
having high Income in 'Academic degree'
Multivariate Analysis for Credit Amount vs Education vs Occupation
Multivariate Analysis for Credit Amount vs
Education vs Occupation
 'Secondary/secondary special' having more outliers.
 'higher education' client having high credits.
Multivariate Analysis for Income Amount vs Education vs Occupation
Multivariate Analysis for Income Amount vs
Education vs Occupation
 'secondary/secondary special' having more outliers
 All Category of Education almost equal in Income.
Multivariate Analysis for Credit Amount vs Education vs Occupation
Multivariate Analysis for Credit Amount vs
Education vs Occupation

 ‘higher education' have more outlier.


 'academic degree' client having high Credit.
Multivariate Analysis for Income Amount vs Education vs Occupation
Multivariate Analysis for Income Amount vs
Education vs Occupation

 'higher education' and 'secondary/secondary special' having


more outlier than others.
 'Academic degree' client having high Income.
Analysis for contract status with loan purposes

 'repair', 'others' and 'urgent need' having high percentage of approval


and refused
Analysis for contract status with Occupation

 'laborers' are having high percentage of approval.


 'IT staff' are having low percentage of approval and unused offer.
Analysis for contract status with Income Type

 'Maternity leave' client are having low percentage of approval.


 'working' client are having high in approved.
Analysis for cash loan purpose with Target
Analysis for cash loan purpose with Target

 Loan purposes with 'Repairs' are facing more difficulites in


payment
 They are 'Buying a garage', 'Business developemt', 'Buying
land', 'Buying a new car' and 'Education' Hence we can focus
on these purposes for which the client is having for minimal
payment difficulties.
Analysis for Occupation with Target

 Occupation with laborers are high in payment difficulties.


 Occupation with IT staff are low in payment difficulties
Analysis for Income Type with Target
Analysis for Income Type with Target

 'Unemployed' and 'Maternity leave' are having low


percentage in payment difficulties
 client with working occupation have high percentage of
payment difficulties.
Analysis for Previous Credit amount vs cash Loan Purpose by
Income Type
Analysis for Previous Credit amount vs cash Loan
Purpose by Income Type
 previous credit amount of Loan purposes 'Buying a home‘ ,
'Buying a land‘ ,'Buying a new car' and 'Building a house' is
higher
Analysis for Income amount vs cash Loan Purpose by Income
type
Analysis for Income amount vs cash Loan
Purpose by Income type
 'repairs' and 'urgent needs' are having more outliers.
 Income amount of loan purpose ‘ bussiness development'
client is higher.
Analysis for Previous Credit amount vs cash Loan Purpose by
Target
Analysis for Previous Credit amount vs cash Loan
Purpose by Target
 Previous credit amount with loan purpose 'buying holiday
home/land' client is higher.
Analysis for Income amount vs cash Loan Purpose by
Target
Analysis for Income amount vs cash Loan
Purpose by Target
 Income amount with loan purpose like ‘ Bussiness
development' are higher.
Analysis for Prev Credit amount vs Housing type
Analysis for Prev Credit amount vs Housing type

 previous credit amount with housing type co-op apartment


and office apartment are higher in defaulters. previous credit
amount with housing type 'with parants' are low in defaulters
Analysis for Income amount vs Housing type
Analysis for Income amount vs Housing type
 income amount with housing type co-op apartment are
higher in defaulters.
 income amount with housing type 'with parents' are low in
defaulters
Conclusion
 Bank should focus on Client with not having Realty. Client currently lived
with parent, Municipal apartment ,office apartment for successful payment.
 Bank should avoid client who lived in co-op-apartment they have high amount
of difficulties in payment.
 Bank should focus on client loan purpose ‘repair’ , ‘business development’.
 In ‘repair’ and ‘buisness development’ loans having high no of rejection and
canceled.
 Banks should focus more on contract type ‘Student’ ,’pensioner’ and
‘Businessman’ with housing ‘type other than ‘Co-op apartment’ for successful
payments.
 Bank should avoid client with less income and unsucessful payment.
 Bank give loans to client with previous loan approved with sucessful payment.
THANK YOU

You might also like