The document discusses a loan prediction system project that aims to automate the loan approval process and make it more efficient. A machine learning model will be developed using datasets to predict if borrowers should be approved for loans based on variables like income, employment status, etc. The proposed system will have a web interface. It will be implemented in two phases, with the front end collecting user inputs and the back end applying a logistic regression model trained on past data to make predictions. The model aims to accurately predict loan eligibility without manual reviews of each applicant.
2. Introduction
In thispresentation,wearegoing to introducethe basicunderstandingofaremini
project.
Thefollowingpresentationincludestheproceduresstartingfromprojectidea,
abstract,scopeofproject,literaturereview, Softwareandhardwarerequirements,
comparativestudy,proposedmodel,conclusionandfuturescope.
3. Eachoneof usget’s thistypeofmessages onein a while
This messages areautogeneratedandthe amountin themessage is alsoauto
generated.
The amountandyoureligibility is calculatedusing machinelearning.
4. Nowadays, sanctioning of loans has become a significant function of the financial
institutions/banking sector.
While sanctioning a loan, the lender needs to have an assurance of earning their money
back along with interest. Thus, identifying the creditworthiness of an individual/an
organization is highly important before sanctioning the loan. In this project, we focus
mainly on monetary loans.
The project aims to thoroughly verify the borrower and perform a background check
based on several variables like gender, income, employment status, etc., to ensure
whether the borrower is creditworthy and can be sanctioned the loan or not.
5. Abstract
Theproject aims at automating the procedure of loan for each
customer, thus, helping in reducing the time and energyand
making the process more efficient.
Theproposed model will use datasets to train itself and giveus
more accurateloan predictions each andeverytime.
Theproject will have a appropriate web page for the customers to
use it on browsers.
6. Problem statement
Why do we need loan predictor?
Whilesanctioningmonetaryloans,banks,andotherlendingauthoritiesneed
tohavea suretyofgettingtheirmoneyback alongwithintereston it.
So,theyneedtoknowthecredibilityoftheborrowerbeforelendingthe
money.Forthis,thelendingauthoritiesneedtoverifythebackgroundand
credibilityoftheborrowerthoroughly.
However,goingthroughseveralvariablesandfactorsmanuallyforevery
borrowerisatime takingprocessandishighlyinefficient.
7. Literature review
Regina EsiTurkson,Edward YeallakuorBaagyere, GideonEvansWenya [1] in their paper have delvedinto the variousmachine learning modelsthat can be
deployedin order to predictthe loaneligibilityofan applicant. They have employed 15 differentlearningalgorithms on the data setin order to determine
which algorithmsare thebestfit for studyingbank credit data sets.
AshleshaVaidya [2] talksabout thetechnical world shiftingrapidly towards completeautomation, thesignificance of automation,and therole ofArtificial
IntelligenceandMachine Learningin it. Usingloan eligibilitypredictionas an example, the author explainshow logisticregressioncan be usedtodesigna
Machine Learningmodel that takes decisionsbased on severalvariablessuch as gender,income, employmentstatus, dependents, etc. to givea final resultas
towhether the borrower in questionis eligiblefor a loan or not
Mohammad Ahmad Sheikh,Amit Kumar Goel, Tapas Kumar [3], throughtheir paper have emphasizedhow theLogisticRegressionModelis an efficientand
effectiveway topredictthe creditworthinessof a borrower. Theauthors useseveralvariablessuch as businessvalue,assets of theapplicant, individualincome
oftheapplicant, etc. to betrained and testedin theLogisticRegressionModel.Theycompared theseresultstotheresults givenby thebanks’systemswhich
mainly focuson theaccount information (which showsthewealth of theapplicant)
8. Requirements.
Hardware
4GB Ram
Inteli3 processor
Internet
Data
Kaggle
Software
Python with anaconda.
Jupyternotebook.
HTML,CSS, JS
11. Conclusion and Future scope.
Ourmodel gives a solution by constructing a Machine Learning based Logistic Regression
model. It is inclusive of all the parameters needed to evaluate the creditworthiness of a
client.
The model is trained to produce results with satisfactory accuracy, after which it produces
accurate results as to whether a borrower should be lent money or not without any
tedious manual work.
12. Reference.
[1].Turkson, Regina Esi, EdwardYeallakuor Baagyere, andGideon Evans Wenya. "A machine learning approach for
predicting bank credit worthiness." 2016 ThirdInternational Conference on Artificial Intelligence andPattern Recognition
(AIPR). IEEE,2016.
[2].Vaidya, Ashlesha. "Predictive andprobabilistic approach using logistic regression: application toprediction ofloan
approval."2017 8th International Conference on Computing, Communication, andNetworking Technologies (ICCCNT). IEEE,
2017.
[3].Sheikh, Mohammad Ahmad, Amit KumarGoel, andTapas Kumar."An Approach forPrediction ofLoan Approvalusing
Machine Learning Algorithm." 2020 International Conference on Electronics andSustainable Communication Systems
(ICESC).IEEE,2020.