December 2019
Intermediate to advanced
368 pages
11h 10m
English
After all agents' evaluation records are collected during the neuroevolutionary process, we are interested in visualizing the recorded data to get insights into performance. The visualization should include the final positions of all solver agents and allow to set the threshold value for the fitness of the species to control what species will be added to the corresponding plot. We decided to present the collected agent records in two plots drawn one above the other. The top plot is for the agent records that belong to the species with a fitness score greater than or equal to the specified fitness threshold, and the bottom plot is for the rest of the records.
The visualization of the agent records is implemented ...
Read now
Unlock full access