This project shows:
- ability of writing solid,structured Python code
- ability of using existing utilities(libraries) for processing and analyzing data.
- analytical and data pre-processing skills
Project is mainly focused on data pre-processing. I showed my techniques in dealing with most common problems with data:
- missing data
- duplicate
- abnormal values.
Project also contains NLP part, where I show the algorithms that I have learned. Moreover, I have demonstrated my ability to effectively present the results to the team by visualizing and summarizing the insights that I have found through out the research.
Data input from the bank - statistics on customers' ability to pay. Each row of the table provides information about the client of the bank: the number of children in the family, employment history, age, education, marital status, gender, type of employment, whether customer had any debts, monthly income, and the purpose of obtaining credit.
| Column | Description |
|---|---|
| children | number of childern in family |
| days_employed | record of employment in days |
| dob_years | customer age in years |
| education | customer education level |
| education_id | customer education level id |
| family_status | family status |
| family_status_id | family status id |
| gender | gender |
| income_type | type of employment |
| debt | whether a customer was in debt |
| total_income | monthly total income in Rubles |
| purpose | purpose to get a loan |
Customer - bank. We need to answer the question: whether the family status and number of children will have an impact on the fact of repayment of the loan on time. Conduct a research on data provided by the bank. The results of the research will be considerer in building a credit score model that assesses the potential borrower's ability to repay the loan to the bank.
pandas numpy matplotlib seaborn nltk missingno