Scribd Logo
Upload

Welcome to Scribd!

  • EDA Assignment

    Uploaded by

    Mallireddy Srinitha
    181 views19 pages
    This document describes an exploratory data analysis case study on loan application data for a consumer finance company. The objectives are to identify patterns that indicate if an applicant is likely to repay a loan or default, in order to minimize losses. The data is cleaned by dropping columns with many missing values, imputing missing values, and changing data types. Outliers are identified through boxplots. Target variables are analyzed, showing imbalance between default and non-default applicants. Bivariate analysis is done to understand differences between defaulters and non-defaulters based on variables like gender, occupation, credit amount, and income. The conclusion recommends limiting credit amounts for male drivers/laborers and female sales/labor applicants who tend to have higher

    Original Description:

    Original Title

    EDA Assignment PPT

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Report this Document
    This document describes an exploratory data analysis case study on loan application data for a consumer finance company. The objectives are to identify patterns that indicate if an applicant is likely to repay a loan or default, in order to minimize losses. The data is cleaned by dropping columns with many missing values, imputing missing values, and changing data types. Outliers are identified through boxplots. Target variables are analyzed, showing imbalance between default and non-default applicants. Bivariate analysis is done to understand differences between defaulters and non-defaulters based on variables like gender, occupation, credit amount, and income. The conclusion recommends limiting credit amounts for male drivers/laborers and female sales/labor applicants who tend to have higher

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Flag for inappropriate content
    Download as pdf or txt
    181 views19 pages

    EDA Assignment

    Uploaded by

    Mallireddy Srinitha
    This document describes an exploratory data analysis case study on loan application data for a consumer finance company. The objectives are to identify patterns that indicate if an applicant is likely to repay a loan or default, in order to minimize losses. The data is cleaned by dropping columns with many missing values, imputing missing values, and changing data types. Outliers are identified through boxplots. Target variables are analyzed, showing imbalance between default and non-default applicants. Bivariate analysis is done to understand differences between defaulters and non-defaulters based on variables like gender, occupation, credit amount, and income. The conclusion recommends limiting credit amounts for male drivers/laborers and female sales/labor applicants who tend to have higher

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Flag for inappropriate content
    Download as pdf or txt
    of 19

    Exploratory Data Analysis

    Name:Mallireddy Srinitha
    Problem-Statement

    This case study aims to give an idea of applying EDA in a real


    business scenario. In this case study, apart from applying the
    techniques that you have learnt in the EDA module, we will
    also develop a basic understanding of risk analytics in banking
    and financial services and understand how data is used to
    minimise the risk of losing money while lending to customers.
    Business-Understanding

    The loan providing companies find it hard to give loans to the people due to their
    insufficient or non-existent credit history. Because of that, some consumers use it as
    their advantage by becoming a defaulter. Suppose you work for a consumer finance
    company which specialises in lending various types of loans to urban customers. You
    have to use EDA to analyse the patterns present in the data. This will ensure that the
    applicants are capable of repaying the loan are not rejected.
    When the company receives a loan application, the company has to decide for loan
    approval based on the applicant’s profile. Two types of risks are associated with the
    bank’s decision:
    If the customer is likely to repay the loan, then not approving the loan results in a loss of
    business to the company
    If the customer is not likely to repay the loan, i.e. he/she is likely to default, then
    approving the loan may lead to a financial loss for the company.
    Business-Objectives

    This case study aims to identify patterns which indicate if a client has
    difficulty paying their instalments which may be used for taking actions
    such as denying the loan, reducing the amount of loan, lending (to risky
    applicants) at a higher interest rate, etc. This will ensure that the
    consumers capable of repaying the loan are not rejected. Identification
    of such applicants using EDA is the aim of this case study.
    Specifications of Application_Data
    ● Shape: (30755,122)
    ● It is a combination of Numerical and categorical variable
    columns
    ● Described the dataset and found the mean, standard
    deviation, minimum value, maximum value, 25%, 50% and
    75% values of each column
    Missing Values in Application_Data
    ● Checked the missing values percentage of each column in
    this data frame
    ● Drop columns with 45% or more Missing values
    After dropping the columns we are left with 73 columns.
    Impute missing values for Numerical Variables

    checking for "AMT_REQ_CREDIT_BUREAU_DAY"


    From this Boxplot we can
    see that there is no outlier and
    value seems to be continous
    so we can impute nan values
    with the mean which is "0.007000"
    checking for "AMT_REQ_CREDIT_BUREAU_WEEK"
    From this Boxplot we can see that there is no outlier and value seems to
    be continous so we can impute nan values
    with the mean which is "0.034362".
    checking for "AMT_REQ_CREDIT_BUREAU_QRT"
    From this Boxplot we can see that there is outlier at "19.0" and "261.0"
    as maximum values are at "0.0" so we can impute nan values with the
    median
    Changing the Data Types
    Changed the datatype of column "DAYS_REGISTRATION" from Float to Int as
    "DAYS_REGISTRATION" contains no. of days which should be integer.
    Changed the datatype of column "AMT_REQ_CREDIT_BUREAU_HOUR" from Float to
    Int as "AMT_REQ_CREDIT_BUREAU_HOUR" contains no. of observation of client's
    social surroundings which should be integer.
    Changed the datatype of "AMT_REQ_CREDIT_BUREAU_DAY” from Float to Int as
    "AMT_REQ_CREDIT_BUREAU_DAY” contains number of enquiries to Credit Bureau
    about the client one day before application which should be an integer value.
    Changed the datatype of AMT_REQ_CREDIT_BUREAU_MON from Float to Int as
    AMT_REQ_CREDIT_BUREAU_MON contains number of enquiries to Credit Bureau
    about the client one month before application which should be an integer value.
    Like this I have changed upto 10 columns.
    Identifying Outliers

    From this figure we can see that except AMT_REQ_CREDIT_BUREAU_QRT


    other columns have continuous values so these cannot be considered as
    outliers.
    From this boxplot we can cleary see that except
    OBS_30_CNT_SOCIAL_CIRCLE and OBS_60_CNT_SOCIAL_CIRCLE other
    columns have continuous values so these cannot be condidered as
    outliers.
    Outlier analysis for AMT_INCOME_TOTAL variable

    There is no huge difference between median and mean


    There is difference between max and 99th and 95th percentile so it has outlier.
    From the Boxplot shown beside, we see that there is outlier.
    Outlier analysis for CNT_CHILDREN variable
    There is no huge difference between median and mean.
    There is huge difference between max and 99th and 95th percentile as a person cannot have 19 children
    as it can be seen that the 99percentile a person has children '3' children so this is outlier.
    From the Boxplot shown beside, we can see that there is outlier.
    Treatment of Outliers By IQR method
    Binning the Continuous variables

    Binning for DAYS_BIRTH Binning for CNT_FAM_MEMBERS

    Here we are dividing DAYS_BIRTH Here we are dividing


    CNT_FAM_MEMBERS column in
    column in different categories of
    different categories of
    Young, Teenage, Old.
    small,medium,large,veryLarge
    We find below observations: Families.

    Young: 225337 We find below observations:


    Small: 80776
    Teenage: 52806
    Medium: 518
    Old: 29368
    Large: 7
    Very Large: 4
    Analysis Of TARGET Variables

    From this below Pie chart we can see the Imbalance between target type
    1 and 0
    ratio of type 1 is 8.07 and type 0 is 91.92
    I have merged the application_data and previous_data datasets and
    created a new dataset as final_data whose shape is (1413701, 110)
    Data divided the data set into 2 subsets based on Target variable Target=0 and
    Target=1
    Looking to the percent of defaulted credits, females have a higher chance of not
    returning their loans.
    Doing Univariate Analysis on Combined data for default customers.
    Doing Univariate Analysis on Combined data for non default customers
    Doing Bivariate Analysis on Combined data for default customers.
    Doing Bivariate Analysis on Combined data for non default customers
    Final Conclusion

    Most of the Male loan applicants are drivers and laborers , and have
    more credit amount and lesser income, so we should limit down the
    credit amount to these kind of clients.
    Most of the Female loan applicants are sales staff and laborers, and
    have more credit amount and lesser income, so we should limit down
    the credit amount to these kind of clients

    You might also like