Rory Mitchell

@Noxoomo @annaveronika if you want to further comment please go ahead, I will be redoing this benchmark next week for a paper submission.

@KruchDmitriy Thanks for your response, I will look at adding some extra datasets.

Not sure what the minimum is, some of the datasets are fairly large. I wouldn't be running it on a laptop anyway as the GPUs are not very powerful.

I was hesitant to set any parameters away from their default values but I think I will do this for the next run.

Try updating the xgboost commit to latest. I would expect to see considerable improvement.

Hmmm strange. My benchmarks show xgboost GPU to be outperforming LightGBM by quite a bit on similar sized datasets. Some factors that could be influencing this: - nrounds = 100...

@Laurae2 good to know. I do have plans to improve this in future. Does a similar problem occur with the CPU algorithm?

You might be interested in cuml: https://medium.com/rapids-ai/accelerating-random-forests-up-to-45x-using-cuml-dfb782a31bea

How many columns in your data set? Is it sparse? If so what is the percentage of missing values.

Will try to go through this in detail when I get time.