Analysis of Loan Data
Analysis of Loan data and relationship with various factors
Introduction
As we all know the history of loans as old as the history of money. Earlier there used to be different mechanism of lending money and recovering it. In simple terms it was the process in which the people who have more money than they required used to give money to people who didn’t had enough. Over the years with the evolution of economics the loan process became extremely important for the people who made business out of it. They used to give loans to people who needed it. But there was always a risk of person defaulting on the loan. For this reason before giving the loan the companies analyses various factors such as the credit history of the borrower, loan period, interest rates, income source etc. in order to prevent any default. In this assignment we are trying to find relation between the interest rates and the various factors like amount, loan length, debt to income ratio, monthly income, FICO score etc.
Methods
Data Collection
The data was collected from the link https://spark-public.s3.amazonaws.com/dataanalysis/loansData.csv provided on the coursera page. The data was downloaded on 14th February 2013 using R software
Exploratory Analysis
Exploratory analysis on the data was done by examining table and plotting the data. The exploratory analysis was used to clean the data and determine factors to be used for the linear regression model. The cleaning of data involved removing inconsistent metrics like year/years, removing percentage signs an converting from factors to numeric for the purpose of regression analysis.
Statistical Modelling
A standard model of multiple linear regressions was built using the R software to check and determine the relationship between the outcome variable and the various factors. Coefficients were calculated and the significance was checked using the P value. The R square value and ANOVA table construction helped in interpreting the result.
The
Introduction
As we all know the history of loans as old as the history of money. Earlier there used to be different mechanism of lending money and recovering it. In simple terms it was the process in which the people who have more money than they required used to give money to people who didn’t had enough. Over the years with the evolution of economics the loan process became extremely important for the people who made business out of it. They used to give loans to people who needed it. But there was always a risk of person defaulting on the loan. For this reason before giving the loan the companies analyses various factors such as the credit history of the borrower, loan period, interest rates, income source etc. in order to prevent any default. In this assignment we are trying to find relation between the interest rates and the various factors like amount, loan length, debt to income ratio, monthly income, FICO score etc.
Methods
Data Collection
The data was collected from the link https://spark-public.s3.amazonaws.com/dataanalysis/loansData.csv provided on the coursera page. The data was downloaded on 14th February 2013 using R software
Exploratory Analysis
Exploratory analysis on the data was done by examining table and plotting the data. The exploratory analysis was used to clean the data and determine factors to be used for the linear regression model. The cleaning of data involved removing inconsistent metrics like year/years, removing percentage signs an converting from factors to numeric for the purpose of regression analysis.
Statistical Modelling
A standard model of multiple linear regressions was built using the R software to check and determine the relationship between the outcome variable and the various factors. Coefficients were calculated and the significance was checked using the P value. The R square value and ANOVA table construction helped in interpreting the result.
The
References: 1. Coursera Data Analysis Course page (https://class.coursera.org/dataanalysis-001/lecture/index) 2. R Tutorial Website (http://www.r-tutor.com) 3. Business Statistics by Levin and Rubin, Pearson Publications