INSPIRATION
Founders struggle to decide where to launch due to many competing regional factors. Key considerations include labour availability, industry density, workforce size, and local competition. We aimed to transform static economic data into an interactive, ML-driven decision engine.
WHAT REGIONMATCH DOES
Converts raw regional statistics into actionable compatibility scores. Uses a trained ML model plus economic indicators to rank UK local authorities. Visualises results in a ranked output table and interactive 3D hex-map.
How we built it
Initial stage involved choosing parameters for the model for which we then found open source data. We then processed and standardized the data, made and trained the model across epochs and then created frontend componets to back it up.
whats next
At the start we were trying to implement an employee side of the web app as well so we may implement a system where instead of helping business owners it could help people looking for jons in the same way.
what we learned
Searching for data could be tedious and time consuming so finding accurate data for the right thing in a short amount of time is insanely essential. More importantly we also learned how to create a proper machine learning models and how exactly the data given to these machines effect the output.
Accomplishments
By making a machine learning module from scratch and using/finding datasets from all over the internet including reddit we even impressed ourselves. We're really proud of being able to finalize a working model for such an important cause like opening businesses.
challenges
The biggest challenge was definitely acquiring the data. Finding the right data for the right output was quite puzzling, there were too much websites and different sources to go through before we eventually landed on proper datasets that helped us in the best way possible
Log in or sign up for Devpost to join the conversation.