Project Abstract:
Financial struggle is always an issue for people and getting loans on the other hand requires good credit history. And the population who struggle to maintain optimum credit history face issues getting their loan approved. And often, corrupt financial lenders take an advantage of this population misleading them in many ways. Home credit Default risk helps people broaden the financial course of action for the undercapitalized population providing a safe and secured borrowing experience. To ensure that this neglected demographic receives a favorable loan,In our experience, Home Credit takes use of a range of other data sources, including telecom and social media. Transactional data—to forecast their clients' repayment ability for the sake of this, we will be utilizing Home Credit Default Risk in this project. While Home Credit is now employing a variety of statistical and machine learning methodologies to make decisions,In order to make these forecasts, we want to leverage past loan application data in this research.We utilized prior loan application data to forecast whether an applicant will be able to repay a loan in this research. This is an example of a standard supervised classification problem. As part of this project, we leverage kaggle datasets to do exploratory data analysis, develop machine learning pipelines, and assess models across many evaluation criteria to deploy a model. In this project, we attempted to accurately identify whether a person is suitable for a house loan using different classic machine learning algorithms like Logistic Regression, Random Forest Classifier, XGBoost, and others. we put in place a deep learning model. Using Pytorch, we created a binary classification Machine learning model. We defined a model, trained it, and assessed it.