Skip to content

Latest commit

 

History

History
37 lines (31 loc) · 1.22 KB

README.md

File metadata and controls

37 lines (31 loc) · 1.22 KB

vaeedaq

Variational AutoEncoder with Population Queue and AVS Copyright Sourodeep Bhattacharjee and Robin Gras, 2019.

To execute the code please enter the following from command prompt console or terminal: Python Main.py

The parameters of the algorithm are in Main.py and described as follows:

######################## PAREMETERS OF VAEEDA Q ######################## population = [] populationHistory = [] N = 1290 # Population Size n = 40 # Problem Size percentParent = 0.5 # Percentage of Parents to Keep in the next generation maxIter = 2000 # Maximum Number of EDA iteration to continue for tournamentSize = 2 # Size of Tournament tournamentProb = 0.8 # Probability of picking best in the tournament fitnessFunc = 1 # 0 - One Max #

headLess = 0 # Graphs are plotted when headless if off(0) , otherwise run in console supressConsole = 0 # Verbose detailed output is shown when set to 0 NKLanscapeTest=0 # Turn this to 1 to test on predefined NK Landscapes NKkvalue=6 # Value of K to use, N is taken from above N

######################## Fitness Function List ######################## 0 - One Max 1 - Trap 5 2 - Trap 7 3 - Trap 9 4 - Trap 11 5 - Trap 13