Credit Risk Analysis for Credit Rating¶

1) Describes and motivates a real-world problem where data science may provide helpful insights. Your description should be easily understood by a casual reader and include citations to motivating sources or relevant information¶

The problem that I am currently looking at is the credit score rating. Credit scores are very important because it can control factors such as buying houses, taking out a loan, and employment. With the advancement of AI in the financial service industry, it can offer an unique opportunitie to imporve the fairness in the credit scoring. Scalable credit risk rating can be made possible by methods and algorithms based on machine learning and data science. The scalability can be built by data engineers to boost platform capacity and Data scientists can adjust and fit models to raise credit scores for the system and the individual users.

Sources used: https://www.theregreview.org/2022/06/07/moss-new-approach-to-regulating-credit-scoring-ai/#:~:text=Credit%20scores%20can%20control%20housing,deepen%20the%20impact%20of%20bias.

https://www.theregreview.org/2022/06/07/moss-new-approach-to-regulating-credit-scoring-ai/#:~:text=Credit%20scores%20can%20control%20housing,deepen%20the%20impact%20of%20bias.

2) Explicitly load and show your dataset. Provide a data dictionary which explains the meaning of each feature present. Demonstrate that this data is sufficient to make progress on your real-world problem described above.¶

Variable name: Credit Scores Data type: Numerical Data Description: Given with a person's credit related information, we will be building a machine learning model that can classify the credit score Scale: 300 - 850 Intepretation: higher values on the scale indicate to lower risk and a higher credit

Variable name: Age Data Type: Numerical Data Description: this will be used to determine the relation between the age and credits Scale: Number of Years Range: positive integer value Intepretation: the older the person this, the higher the credibility might be due to their job experience

Variable name: Income Data Type: numerical Data Description: the annual income per person Scale: US dollars Intepretation: this will be used to determine the relation with a higher income which might mean a greater financial stability and therefore might have a higher credit score

Variable name: Number of Bank Accounts & number of credit cards Data Type: Numerical Data Description: this is the total number of bank accounts a person would have Scale: by count Range: positive values Intepretation: number of bank accounts could affect credit rating due to more credit cards meaning more money to pay back. Good thing or bad thing - depending on how much they are able to pay back.

Variable names: Interest Rate Data Type: numerical data Description: the interest rate on the credit account Scale: percentage Range: positive Intepretation: higher interest rate may result in making more consistent payments which might raise higher risk and potential higher cost

Variable names: Number of loans Data Type: numerical data Description: the number of loans a person may have Scale: count Range: positive integer Intepretation: the higher the loan, the higher the credit. It may indicate that one has the ability to pay off loans within a period of time

Variable names: type of loan Data Type: categorical Description: the type of loan a person may have Scale: depends Range: depends Intepretation: the level of loan may indicate one's financial stability

In [7]:

import pandas as pd

pdDf = pd.read_csv('test.csv')
pdDf.head()

Out[7]:

	ID	Customer_ID	Month	Name	Age	SSN	Occupation	Annual_Income	Monthly_Inhand_Salary	Num_Bank_Accounts	...	Num_Credit_Inquiries	Credit_Mix	Outstanding_Debt	Credit_Utilization_Ratio	Credit_History_Age	Payment_of_Min_Amount	Total_EMI_per_month	Amount_invested_monthly	Payment_Behaviour	Monthly_Balance
0	0x160a	CUS_0xd40	September	Aaron Maashoh	23	821-00-0265	Scientist	19114.12	1824.843333	3	...	2022.0	Good	809.98	35.030402	22 Years and 9 Months	No	49.574949	236.64268203272135	Low_spent_Small_value_payments	186.26670208571772
1	0x160b	CUS_0xd40	October	Aaron Maashoh	24	821-00-0265	Scientist	19114.12	1824.843333	3	...	4.0	Good	809.98	33.053114	22 Years and 10 Months	No	49.574949	21.465380264657146	High_spent_Medium_value_payments	361.44400385378196
2	0x160c	CUS_0xd40	November	Aaron Maashoh	24	821-00-0265	Scientist	19114.12	1824.843333	3	...	4.0	Good	809.98	33.811894	NaN	No	49.574949	148.23393788500925	Low_spent_Medium_value_payments	264.67544623342997
3	0x160d	CUS_0xd40	December	Aaron Maashoh	24_	821-00-0265	Scientist	19114.12	NaN	3	...	4.0	Good	809.98	32.430559	23 Years and 0 Months	No	49.574949	39.08251089460281	High_spent_Medium_value_payments	343.82687322383634
4	0x1616	CUS_0x21b1	September	Rick Rothackerj	28	004-07-5839	_______	34847.84	3037.986667	2	...	5.0	Good	605.03	25.926822	27 Years and 3 Months	No	18.816215	39.684018417945296	High_spent_Large_value_payments	485.2984336755923

5 rows × 27 columns

3) Write one or two sentences about how the data will be used to solve the problem. Earlier in the semester, we won’t have studied the Machine Learning methods just yet but you should have a general idea of what the ML will set out to do.¶

The credit risk data can be used to evaluate the creditworthiness of borrowers and mamke more informed decisions about credit applciations, interst rates, and credit limits. This helps solve credit rating problems by minimizing the risk of default or non-payment on loans and credit accounts.