Added functionality for producing aggregated plots for a single task in a given experiment.
Here each plotting point for a given algorithm on a particular task will be the mean over all independent experiment runs and the error will be the 95% confidence interval.
Why?
To better match the guideline.
How?
Created plotting and aggregation functions.
Extra
The 95% confidence interval is implemented using the central limit theorem for computational efficiency, but bootstrapping can be used instead.
What?
Added functionality for producing aggregated plots for a single task in a given experiment. Here each plotting point for a given algorithm on a particular task will be the mean over all independent experiment runs and the error will be the 95% confidence interval.
Why?
To better match the guideline.
How?
Created plotting and aggregation functions.
Extra
The 95% confidence interval is implemented using the central limit theorem for computational efficiency, but bootstrapping can be used instead.