Skip to content

Conversation

@gmpgreen
Copy link
Owner

Structures for GA population & simple fitness function.

Note: Created own build script on local machine for building the source files in the project. Have not updated CMAKE with new files.

gmpgreen added 9 commits July 15, 2019 17:31
Structures for GA population & simple fitness function
Minimum viable product, can successfully determine spam messages with
over 88% accuracy if iterations set high enough. (note: currently
training data and test data are the same, must change later)
Fixed memory errors from storing too many local copies of large vectors
-Seperated data used for training (5524 messages) and testing (50 messages)
-Implemented a new random function for generating random numbers used
-Changed weighting of chromosomes during tests and training
Note because seed is now randomized for random numbers, reults will vary between each execution of the program.
-Added finish condition for when there is less than 12 unique chromosomes left
-Updated test data from 50 messages to 100 messages for more consistant results
-minor comments and tweaks
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants