Conventional credit scoring methods like credit history, collateral, or any prior accounts might not give every person an equal opportunity to establish credit. As a result, a small number of people can access basic banking services like opening a bank account and acquiring a loan. Using the pre-existing datasets, we can understand the track record of banks and the reason why conventional credit scoring is needed. Come up with an AI based solution to perform credit scoring for people who don’t have any previous loans. This can be done by checking some other billing cycles they’ve been a part of and creating a scoring metrics on the basis of that.
In our commitment to bridging financial disparities, our project focuses on empowering those often excluded from basic banking services, particularly in obtaining loans without an established credit score.
Our innovative AI solution delves into the wealth of historic customer behavior data. By leveraging advanced analytics, we can predict the creditworthiness of new customers lacking a loan history. This predictive capability allows us to discern the level of risk associated with each individual.
At the heart of our solution lies a foundational Logistic Regression model. This statistical technique, coupled with machine learning principles, becomes a powerful tool in evaluating credit risk. Through meticulous data preprocessing and cleaning, we refine raw datasets, ensuring the model's accuracy and reliability.
Kaggle Dataset Exploration: Dataset Used
To empower our AI with robust insights, we tapped into Kaggle, a premier platform for datasets. Our team explored various correlations and attribute combinations, gaining a nuanced understanding of the data landscape. This exploration laid the groundwork for constructing a solution that not only predicts but also adapts to dynamic financial scenarios.
Building on the power of Flask, we crafted an intuitive and engaging web platform. Our user interface is designed to be user-friendly, ensuring a seamless experience for individuals seeking financial solutions.
The frontend of our application is a testament to our commitment to accessibility. Through thoughtful design and meticulous implementation, we've made interacting with complex financial algorithms a breeze for users with varying levels of technical expertise.
Taking the leap from development to deployment, we chose Render as our hosting platform. Render's scalability and ease of use provided a perfect environment for our Flask app. This transition ensures that our AI-driven credit scoring is not confined to labs but accessible to anyone, anywhere.
- Ayushi Dubey (Email Id, GitHub Profile Link)
- Monya Mehta (Email Id, GitHub Profile Link)
- Manasvi Jindal (Email Id, GitHub Profile Link)