Introduction (20 points) Domain and Motivation Our project delves into the domain of personal finance, focusing on a pressing issue many consumers face today: credit card debt consolidation and balance transfers. The motivation behind this project stems from the complexities and challenges consumers encounter when navigating the myriad of options available for managing and consolidating credit card debts. This endeavor aims to streamline this procedure, transforming it to be more transparent and user-friendly, regardless of their level of financial expertise.
Terminology/Jargon Firebase - is a popular development platform offered by Google and is widely used in building mobile and web applications. CSS - Cascading style sheets are the cornerstone technology of the web, alongside HTML and JavaScript. It is used for the styling and layout of web pages. Javascript is a powerful and versatile programming language widely used in web development. SQLite3 - a lightweight, usually embedded database that does not require a separate server process. Machine Learning is a branch of artificial intelligence that focuses on building systems that are capable of learning and making decisions given different kinds of data. Random Forest - Machine learning algorithm that is used for classification and regression. Django - is an open-sourced web framework written in Python, designed to facilitate the rapid development of web applications.
History
Historically, the implementation of debt consolidation and balance transfer has always been a lengthy procedure, copious amounts of documentation, and requires manual evaluation of financial instruments. On the contrary, although these procedures have been significantly optimized since the inception of financial technology, a noticeable gap persists regarding the availability of a cohesive platform that streamlines comparisons and facilitates decision-making regarding individual financial profiles.
Objectives
Our project's primary objective revolves around developing a user-friendly web application to streamline the comparison of debt consolidation and balance transfer options across various credit cards. Our overarching goals include providing consumers a platform to evaluate credit card choices tailored to their individual profiles and financial objectives. Specifically, we aim to implement a Django front-end landing page, integrate percentages for a machine learning algorithm to analyze credit card options, and develop a web form to collect user data. We've successfully addressed challenges related to the Django project's logic, machine learning algorithm implementation, and web form integration.
However, obstacles arose in obtaining accurate data from credit card companies through APIs, leading us to pivot our strategy. Our revised plan involves using sample data from other websites for testing, focusing on algorithm training and validation to reduce external dependencies. The success of our project hinges on overcoming risks associated with ensuring accurate database collection of user data, managing project timelines amidst other class commitments, comprehending Django back-end programming intricacies, and presenting the project effectively.
Our project aligns with the current state of financial technology by addressing dynamic challenges in credit card decision-making. Our web application stands out in the era of personalized and data-driven solutions, providing users with a sophisticated tool to compare debt consolidation and balance transfer options based on their unique financial profiles. Integrating a machine learning algorithm reflects a forward-looking approach, leveraging cutting-edge technologies. Our adaptation to using sample data for testing demonstrates a pragmatic response to challenges, contributing to the ongoing evolution of financial technology.
Methodology
Machine Learning Model A 4000-row dataset of loan and credit card application data was obtained from Kaggle to train a machine learning model to evaluate the likelihood of loan repayment. The Sklearn Python library was used to import the dataset and split the data into training and evaluation datasets. After the data was split, a Random Forest Regression model was created. The loan status was mutated into binary values of 1 and 0, indicating loan repayment and default, respectively. Loan status was set as the dependent variable, and other indicators, such as assets, income, FICO score, etc., were selected as the independent variables. A Random Forest of 10,000 trees was created and then fit with a mean absolute error of 0.08. With the model fit, predictions could then be made. The model produces a score of between 0 and 1. A score of 0.50 or greater is required for approval, and others are denied. This model proves efficacy providing efficient and accurate loan decision-making.
Results (25 points) Achievements and Quantitative Results
The development of our web application for credit card debt consolidation and balance transfer has achieved several significant milestones. We have successfully implemented a user-friendly interface, making it accessible and intuitive for users of varying levels of tech-savviness. This interface provides a seamless experience, from account registration to viewing and managing debt balances.
One of the project's significant accomplishments is integrating a machine learning algorithm, utilizing users' debt ratios and credit scores to identify and recommend the most beneficial balance transfer options. This feature not only enhances the user experience but also adds a layer of personalization, tailoring recommendations to individual financial situations. Additionally, we included the integration of educational resources on personal finance management to enhance the application's value proposition, which also aligns with our goal of making financial management more accessible to a broader audience.
Finally, we were able to utilize many web tools in our program. Using Django, we used HTML, CSS, JavaScript, and Python to create a functional and visually pleasing platform. Django’s SQLite3 database was used to store the user's responses and pulled from the Python machine learning code. The last tool we implemented was the Firebase authentication for signing into the application. We can pull the user’s name from the login information for a customized dashboard.
Learning Outcomes and Challenges Throughout the project, our team has encountered various challenges and learning opportunities. One significant challenge was the pivot from using APIs from credit card companies to employing sample data for testing and algorithm training. This decision, though necessary, required us to adapt our approach to data acquisition and integration. Our team also navigated through the complexities of Django for both front-end and back-end development. Understanding and effectively utilizing this framework was crucial for integrating different components of our application, especially in ensuring seamless navigation and user data management. Another learning aspect was balancing project management with technical development. This involved aligning project deadlines with class requirements and managing the learning curve associated with new technologies and programming languages.
Conclusion (30 points)
Summary of Findings Our project has successfully demonstrated fintech's potential in revolutionizing how consumers manage credit card debt. Our team's web application achieves its intended functionality and adds value through its user-centric design and innovative features. Ultimately, it effectively bridges the gap in the market for a unified platform that simplifies the process of credit card debt consolidation and balance transfers. The core functionalities of our application, including the machine learning algorithm utilized to generate personalized recommendations, the capability to link and access credit card data, and the capability to commence and monitor balance transfers, have proven to be incredibly influential. These features give users a new level of control and insight into improving their financial health.
Insight/Reflection Our team has acquired invaluable insights into the convergence of finance and technology over the course of this project. Having a better understanding of the criticality of user experience design when developing financial applications, especially the importance of functional, intuitive, and accessible applications that accommodate users with diverse financial expertise. Furthermore, this project demonstrated how critical it is for adaptability and resiliency when it comes to project management. Our team quickly pivoted from the planned infrastructure in light of external constraints, as demonstrated by our transition from utilizing external APIs to employing sample data for development.
Future Developments In the future, there are several avenues for further development and expansion of the project. Our immediate goal is to release the minimum viable product (MVP) and obtain user feedback to improve the application’s features and user-friendliness. Recognizing the increasing use of mobile applications, we also intend to develop a mobile version of our platform, to accommodate the convenience of financial management on the go. Next would be incorporating more diverse financial products, such as loans and mortgages, which could broaden the application's scope and utility. Additionally, advanced machine learning algorithms could be implemented for more nuanced financial recommendations, considering factors like spending habits and financial goals. The successful implementations of APIs can also help users to seamlessly and securely connect to their accounts, as well as exploring the possibility of introducing an innovative feature that enables users to directly close or consolidate cards within the application. The integration of this feature will revolutionize the process of managing debt by providing users with an enhanced convenience in optimizing their credit card accounts. In conclusion, this project has achieved its immediate objectives and paved the way for further exploration in personal finance and fintech, demonstrating our team's ability to overcome challenges and adapt to changing circumstances.