Embedded Files
General Purpose Machine Learning Through Genetic Programming
Harness the power of evolutionary processes to generate optimal programmatic solutions.
FloraML uses evolutionary concepts from biological systems to find optimized solutions to problems. The input is training data and a desired benchmark to optimize for. The output is a population of Turing-complete procedural programs. These programs consume input data and perform operations on it so their outputs meet some criteria.
Unlike neural networks, and other popular ML methods, it is very easy to look into the programs that FloraML generates and understand WHY and HOW it is doing what it does.
It's a simple process...
1. Collect training data (boolean and numerical) and format into CSV files
2. Create/generate initial population of candidate solution programs
3. Measure fitness of the population individuals against the desired performance (e.g. minimize this, maximize that)
4. Select best performing individuals and mutate them to create a new generation of individuals
5. Repeat steps 3 and 4 until one or more individuals (solution programs) meet the desired performance
Market Modeling
Market Modeling
Mathematical Modeling
Mathematical Modeling
Model Optimization
Model Optimization
Questions?
Questions?
Contact FloraML@savoyengineering.com to get more information on the project
Page updated
Google Sites
Report abuse