What it does
Home for your $$$ is able to take in data of other houses such as their bedroom/bathroom count, square footage, the year it was built, and their existing prices. Using that data, it runs it through an algorithm that can predict what a house might cost with features that you input yourself! Data can be input manually by hand or the option of prebuilt csv files.
How I built it
UIPath was utilized for the front end and connecting all components driving the backend. Microsoft Excel was used mainly as a database for storing the large amounts of training data in memory. The functions of Excel were also taken advantage of in UIPath's routines. UIPath handles all instructions, front end prompts, and all read, write, and data manipulation for Excel.
Challenges I ran into
We had a hard time getting the main learning algorithm to work properly. We had to make use of a very extensive Excel formula and go back to Linear Algebra to understand what we were doing completely. Furthermore, getting UIPath to interact with Excel through importing .csv data and making sure we were pointing to the right locations in Excel had proved to be some issues. The biggest challenge we had was learning how to use UIPath as it is something we had never used before, but we have gotten pretty familiar with it.
Accomplishments that I'm proud of
We managed to get UIPath, Excel, and CSV files all to work in harmony. Our biggest accomplishment was getting the full algorithm to interact flawlessly after much trial and error.
What I learned
We learned what and how to use UI Path and how it can drive many parts of a program. We learned how to use it for front end work and also driving back end processes. This project also encouraged me to learn how to use many different Excel functions, practice with Github, and a different way in thinking of programming through flowchart logic.
What's next for Home for your $$$
A feature that we wished we could have implemented was being able to actively scrape data live from a personal home real estate website, getting information about the homes so we can predict house data using the live market as opposed to pre-written data. A web scraper is definitely a future feature.
Log in or sign up for Devpost to join the conversation.