Comparative Analysis of Ml Algorithms & Stream Lit Web Application

Abstract

Prediction of Approval of Consumer Personal Loans Applications using different Machine Learning (ML) Algorithms and convert it into a web app by using stream lit library in a very simple and efficient way. Out of the available ML algorithms a comparative analysis is performed to check which one proves out to be the best possible with maximum accuracy. This project basically provides the assumption or probability of approval or denial of the personal loan any candidate applies for. As soon as the customer provides in his or her details and all domains that are required. Various models are used to test whether the given details of the customer are okay to provide loan or not. If yes, it displays a message" loan approved" and further functionalities can be done. Else it displays a message as "Loan denied". Understanding of Test and Train Data. Implementation of steps like Feature Engineering, Feature Selection, Exploratory Data Analysis (EDA), Model Training etc. Usage of different models to train the data like below. LR, RF, XGB. Compares the accuracy from different models. The customer is asked to fill in the details required whoever wishes to apply for the loan. According to the given data if it’s suitable and all the conditions are satisfied, the user gets a message related approval of the loan else it displays a message of denial.

Topics & Concepts