Hello
I am attempting to apply Genetic algorithm on KDD Cup 99 data set in order to generate genetic rules Currently i have initialized population however i don't know based on what will i calculate the fitness
Any help is appreciated , Thanks in advance
Salam,
you should determine where you want to apply genetic algorithm?
Are you going to employ genetic algorithm to solve feature selection problem or tuning the parameter of existing classification algorithms ?
You can use support-confidence framework to calculate fitness.
I am attempting to apply genetic algorithm on training data to training data set in order to come up with a set of rules which can be applied on testing date set to specify whether a specific record is an attack or not
I am not that mush familiar with genetic algorithm unfortunately however i was able to accomplish the below steps
I was able to read training data set
I initialized a new population
I applied crossover and mutation on the new population
However i don't know based on what should i select the best individual from the new population which fitness function should use ??
Please see my paper on applying optimization algorithm on feature selection.
https://www.researchgate.net/publication/267427043_Bees_Algorithm_for_feature_selection_in_Network_Anomaly_detection
Article Bees Algorithm for feature selection in Network Anomaly detection
Refer following papers...
http://airccse.org/journal/nsa/0312nsa08.pdf
http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.447.3397&rep=rep1&type=pdf
The following paper can give you some insights on which fitness function you could use:
http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.429.3542&rep=rep1&type=pdf
- Badges
- Science topic
- Similar topics
- Engineering
- Computer Engineering