Novice Finance

Financial Empowerment Through A.I.

Comment

Inspiration

In today's technologically advanced society, the business and finance worlds are dominated by wealthy and powerful MNCs with vast capital and technological resources. Lacking similar resources and factors of production, novice entrepreneurs and lower class citizens have little opportunity for economic growth and business development. As such, Novice Finance was constructed with the hope of fulfilling UN's goal #8 by providing efficient financial tool to allow lower class citizens and novice entrepreneurs to attain sustainable economic growth through implementing world-class financial tools

What it does

Novice Finance provides four machine learning modules that act as financial tools for burgeoning entrepreneurs and prospective citizens for their financial goals. Currently, the machine learning modules employed are: Startup Success Rate Estimator, Loan Approval Rate Predictor, Default Chance Calculator, and Customer Churn Forecaster The Startup Success Rate Estimator captures user input on specific information about a startup’s financial and logistic details to predict the chances of that startup’s success. The Loan Approval Rate Predictor utilizes information about an individual’s financial status to determine the chance of that individual’s specific loan amount being approved. The Default Chance Calculator requires financial information about an individual’s previous financial transactions and payments to estimate the chance of credit default for the individual in the near future. The Customer Churn Forecaster captures input on demographic and financial information about a specific client of a business to forecast the chances of that client exiting the business in the future.

How we built it

Novice Finance was built through Visual Studio Community and Razor Pages in ASP.NET Core. On the server side, C# was utilized for the implementation of the four machine learning modules, and on the client side, cshtml was used for capturing user input and the overall web design. ML.NET was employed for the creation of the machine learning module frameworks. Collectively, user input on the client side Razor Page is captured through the cshtml forms, and this information is transferred to the ML.NET modules through the server side C# pages. There, the prediction by the machine learning modules is outputted back to the Razor Pages for the users to view the results of our A.I. powered prediction.

Challenges we ran into

One of the major challenges we ran into was figuring out how to transfer user input from the client side html page to the server side C# page, where the machine learning modules were established. We solved this problem by utilizing cshtml instead of html for the client side Razor Pages. Utilizing cshtml enabled us to establish a direct connection between the client side Razor Pages and the server side C# page, thereby allowing for direct transfer of information between the two sources.

Accomplishments that we're proud of

Producing a live and accessible web application that can be used by millions around the world for all their financial needs is the biggest accomplishment that Novice Finance has attained.

What we learned

One of the major computer science related concepts learned from this project was learning how to construct and implement the ML.NET machine learning modules from Visual Studio Community. The ML.NET model builder provides a simple, yet powerful framework for constructing machine learning modules using C#. Learning to use ML.NET was a significant stepping stone in our computer science careers, as it enables us to effectively implement artificial intelligence into any CS project.

What's next for Novice Finance

Moving forward, we would like to expand the scope of Novice Finance by incorporating machine learning modules for more financial tools. Possible areas of focus would include Financial & Credit scam detection, product demand prediction, assets management, and financial risk estimation. To improve the functionality of our current machine learning modules, we would like to expand the financial databases utilized for the creation of our ML.NET modules, by incorporating financial data from prominent financial institutions. To those ends, we will take necessary steps to obtaining the necessary credentials required, with the hopes that an improved database source would make our prediction more accurate.

Built With

Share this project:

Updates