Vous êtes sur la page 1sur 9
MANAGEMENT PROJECT - PROPOSAL APPLICATION OF ANALYTICAL TECHNIQUES FOR LOAN PREDICTION Kartik Mehta 7/25/16 MANAGEMENT PROJECT

MANAGEMENT PROJECT - PROPOSAL

APPLICATION OF ANALYTICAL TECHNIQUES FOR LOAN PREDICTION

Kartik Mehta

7/25/16

MANAGEMENT PROJECT

Contents

Description of the Project

2

Predictive Modelling in Finance Sector

2

Introduction

3

Overview of Global Analytics Market

3

Need for Predictive Modelling

3

About the Project

3

Primary Objective of Research

4

Proposed Methodology

5

Variables in the Data Set (as provided by Analytics Vidya)

5

Managerial

Implications

6

Limitations

of the study

7

 

8

Description of the Project

Predictive Modelling in Finance Sector

Company wants to automate the loan eligibility process (real time) based on customer detail provided while filling online application form. These details are Gender, Marital Status, Education, Number of Dependents, Income, Loan Amount, Credit History and others. To automate this process, they have given a problem to identify the customer’s segments, those are eligible for loan amount so that they can specifically target these customers.

Introduction

Overview of Global Analytics Market

Analytics market and need are on a rise. Managers no longer depend on their intuition but rather depend on data backed or supported decisions. Global BI, CPM and Analytics applications / performance management software revenues is worth US $ 13 Billion in 2012, a 6.8% rise from 2011. Scope of analytics products has been evolving rapidly and far beyond conventional reporting based on data extracted from core enterprise applications like ERP and CRM.The Global Analytics Market is worth US$ 13 Billion.

Need for Predictive Modelling

Predictive analytics has come of age as a core enterprise practice necessary to sustain competitive advantage. This technology enacts a wholly new phase of enterprise evolution by applying organizational learning, which empowers the business to grow by deploying a unique form of data-driven risk management across multiple fronts

The objective of predictive modelling is to help to manage data and provide insightful results. These insightful results can help the data owner to make statistical based decisions and have data to support the decisions rather than take decisions simply based on intuition.

We need predictive modelling for the purpose to compete as data insights will help us secure a powerful and unique competitive stronghold which shall help companies to grow and enforce business integrity. Slashing costs is also an added advantage of predictive modelling. For instance, it can help company decrease campaign spend by identifying respondents.

About the Project

For this project we have considered data set from data competition. A data competition provides a real time business problem and provides segregated data. The data is segregated into Training Data set and Test Data set. We use our analytical tools and techniques on training data set and use the results obtained and apply it on test data set to measure the robustness of the model. Series of techniques are used to make the model and we select the model which has the least errors or residuals and which provides good degree of accuracy.

The considered data set is a Loan Prediction Problem. The problem is to predict with certain degree of accuracy coupled with least number of residuals or errors if the next person is eligible to receive a loan or not. The data will be cleaned and analyzed using analytical tools. The preferred tools are SAS, XL Miner, Tableau.

Primary Objective of Research

Our objective is to understand the following:

  • 1. Understand the data and learn the application of data manipulation techniques to prepare data for modelling.

  • 2. Learn, identify and apply the modelling techniques which helps us to identify the best model.

Proposed Methodology

Empirical / Qualitative

Empirical

Data collection method

As provided by data competition

Sample size

As provided in the data by data competition

Respondents

As provided by data by data competition

Statistical techniques to be used

Logistic Regression, Decision Tree, Neural Networks

Variables in the Data Set (as provided by Analytics Vidya)

Variable

Description

Loan_ID

 

Gender

Unique Loan ID Male/ Female

Married

Applicant married (Y/N)

Dependents

Number of dependents

Education

Applicant Education (Graduate/ Under Graduate)

Self_Employed

Self-employed (Y/N)

ApplicantIncome

Applicant income

CoapplicantIncome

 

LoanAmount

Coapplicant income Loan amount in thousands

Loan_Amount_Term

Term of loan in months

Credit_History

 

Property_Area

credit history meets guidelines Urban/ Semi Urban/ Rural

Loan_Status

Loan approved (Y/N)

Managerial Implications

The idea of the project is to help corporates manage their data and bring out meaningful insights. The predictive modelling will help the company to automate the

loan eligibility process and get insightful information of their customers thus making it easy for the company to identify which factor plays an important role in deciding the eligibility of the customer for the loan.

Limitations of the study

The limitation of the project is:

Limited to no interaction with the client thus limiting our access or need for further or new variable data.

References

  • 1. Anon., 2016. NASSCOM. [Online] Available at: http://www.nasscom.in/india-%E2%80%93-hub-analytics-products-analytics- product-excellence-matrix [Accessed 18 July 2016].

  • 2. Anon., n.d. Analytics Vidhya. [Online] Available at: http://datahack.analyticsvidhya.com/contest/practice-problem-loan-prediction-iii [Accessed 18 July 2016].