Is the average ranking meaningful since each algorithm is test on different number of datasets ?

[abvesa]

No description provided.

[duncanmcelfresh]

it seems this is a question about our paper and not this repo. good question though - when we show aggregated rankings (min, max, mean, median, ...) we are aggregating over the same group of datasets for each algorithm. so: when we show average rankings for different algorithms, these averages are taken over the exact same datasets.