Scribd Logo
Upload

Welcome to Scribd!

  • 3 views

    t2

    Uploaded by

    kokobhaiya143

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd

    t2

    Uploaded by

    kokobhaiya143
    3 views10 pages

    Copyright

    © © All Rights Reserved

    Available Formats

    DOCX, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd
    Download as docx, pdf, or txt
    3 views10 pages

    t2

    Uploaded by

    kokobhaiya143

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd
    Download as docx, pdf, or txt
    You are on page 1of 10

    Introduction:

    The Loan Approval Classification Dataset is a collection of financial data used to


    predict whether a loan application will be approved or denied. It includes
    information about the applicant's demographics, employment, income, and credit
    history. This dataset is commonly used for training machine learning models to
    automate the loan approval process.

    The objective of the Loan Approval Classification Dataset is to predict whether a


    loan application will be approved or denied based on various financial and
    demographic factors. This dataset is used to train and evaluate machine learning
    models that can automate the loan approval process, making it more efficient and
    accurate.

    1. Handling Missing Values

    The script first looks for missing values in the dataset using some handy R
    functions like is.na() and colSums(). It calculates the total number of
    missing values and even breaks it down column by column. For example:
    Once the missing values are identified, the script offers a couple of solutions:

     It uses na.omit() to completely remove rows with missing values if


    necessary.
     For more refined handling, it fills missing values with things like the
    median of the column. For example:

    So, it makes sure the missing data doesn’t mess up the analysis.

    2. Visualizing Missing Data

    To make things clearer, the script uses graphs to visualize where data is missing.
    This helps in understanding the problem better. It uses two tools:

    1. Naniar Library: With gg_miss_var(), it creates a simple graph to


    show how many missing values each column has.
    2. VIM Library: The aggr() function makes a neat plot, coloring missing
    values in red and complete ones in blue. This is like a visual heatmap of the
    missing data.

    Here’s an example:
    3. Balancing the Dataset

    The dataset might have an imbalance in how many rows belong to each class, for
    example, in loan_status. If one class (like "Approved") has way more rows
    than another (like "Rejected"), it can skew the analysis.

    The script fixes this using the ROSE library. It tries three methods:

    1. Oversampling: Adding more rows to the smaller class.


    2. Undersampling: Removing rows from the larger class.
    3. Combination (Both): A mix of oversampling and undersampling.

    For instance:
    Now, the data is balanced, making the analysis fairer.

    4. Removing Duplicate Data

    Duplicates in a dataset can distort results. The script identifies rows that are
    repeated using the duplicated() function. For example:
    Once duplicates are identified, the script removes them. It can also find duplicates
    based on specific columns, like person_age,person_gender, person_gender and
    person_education.

    This ensures that the data is clean and has no redundancy.

    5. Filtering the Data

    The script demonstrates several ways to filter rows of data based on conditions.
    For example:

     To find only male participants and to filter for people aged between 20 and
    30:
    This is useful when you want to focus on specific segments of the data.

    6. Converting Between Categorical and Numeric Data

    Sometimes, categorical data like "Male" and "Female" needs to be turned into
    numbers (0 and 1) for better analysis. It also turns education levels (like
    "Bachelor", "Master") into numbers:
    This makes the data easier to work with for statistical models.

    7. Normalization
    Normalization is when you scale numeric values so they fit within a specific range,
    like 0 to 1. The script normalizes person_income using min-max scaling:

    This is useful when different attributes have very different scales.

    8. Handling Outliers and Invalid Data

    Outliers, or values that are way too high or low compared to the rest, can mess up
    the analysis. The script identifies outliers using the Interquartile Range (IQR)
    method.
     First, it calculates the lower and upper bounds:

     Then, it replaces outliers with the median income:


    For invalid data (like negative ages), it simply identifies and removes or fixes
    them.

    You might also like