Average of Best Fitness Per Generation#
This section provides a detailed view of how the best individual in each generation evolves across all experiments within a given round. The data is represented in a line chart that displays the mean best fitness per generation.
Each point in the curve represents the average fitness value of the best individual from each experiment for a given generation.
To illustrate how this is computed, consider the table shown below:
Take generation 1 as an example:
Experiment 1: Best fitness = 0.5702
Experiment 2: Best fitness = 0.5853
Experiment 3: Best fitness = 0.7656
Average = (0.5702 + 0.5853 + 0.7656) / 3 = 0.6404
This average value is plotted as the point corresponding to generation 1 on the chart.
The same process is repeated for all 20 generations, producing a smooth curve that allows the user to evaluate:
Convergence speed: How quickly the algorithm approaches optimal values.
Stability: Whether the curve stabilizes or oscillates.
Performance: General fitness trend across time.
Note: This visualization is fundamental to assess the efficiency of parameter settings and identify whether the genetic algorithm is behaving as expected.