Prediction Model - Mixed Martial Arts
Everyone has dreamed of the ability to be able to predict the future. What if it were possible? What if the outcome of a Mixed Martial Arts (MMA) competition could be determined through algorithms, math, and data? We believe that it can.
Objectives
To create an outcome prediction model for the sport of Mixed Martial Arts (MMA). The prediction model would harness historical competitor and competition data to calculate likely outcomes.
To achieve the main objective of this project, several sub-objectives will need to be achieved:
- Develop or modify an existing web scraping application to extract competitor and competition data from web sites.
- Create a custom database to hold the web-scraped data.
- Identify key metrics within the dataset that likely leads to an expected outcome.
- Assign weight values to key metrics (some metrics are more influential to an outcome than others).
- Develop recursive algorithms that identify likely key metrics and weights.
- Develop recursive algorithms that test the prediction model and self-adjust parameter values for prolonged testing runs.
- Explore where machine learning could be implemented to make testing more autonomous
- Create graphical charts showing influential metrics, accuracy percentage of the model between testing runs, etc.lead
- Develop a CLI to be used during development.
- Develop a GUI to display results and to interact with the application (production).
Motivations
The success metric of this project is definable and easily measured - 51% accuracy. Why stop there? Perhaps 75% accuracy is achievable... Maybe 85%. The ramifications of a successful model could prove to be very fruitful, in both academia and the professional world.
Qualifications
Minimum Qualifications:
- Have experience with database
- Experience with recursive algorithms
- Creative thinker
- The ability to recognize patterns and analyze data
None Listed
Details
Project Partner:
James Cole
NDA/IPA:No Agreement Required
Number Groups:1
Project Status:Accepting Applicants
