Alex Rogozhnikov

Hi, Simone. You're doing somewhat strange and expect algorithms to do the things they can't know about. Cross-validation of machine learning is easy when you have some figure of merit...

> Would that be useful? Not really. 1-dimensional discrepancies are not all discrepancies. You can drive 1-dimensional ROC AUCs to 0.5 with max_depth=1, but you'll not cover any non-trivial difference...

If you really want to automate this process, you need to write evaluation function which encounters both steps 1) and 2) mentioned above. E.g. sum over KS(featuture_i) + abs(ROC AUC...

Dear Vincente, there is no difference between single / multidimensional fits. The data you pass should be `numpy.array` or `pandas.DataFrame` of shape `[n_samples, n_features]`. If you want to pass weights,...

Hi Simon, right now we present this as a research, not a production-ready tool. Comparison with other libraries would be nice, but to understand the effect of proposed boosting scheme,...

Hi Hristo, indeed, there is no ground-breaking improvement. However in both cases algorithm was compared on the problems where competitors are very good and using tree hyper-parameters that are known...

@liu044100 feel free to fix it with pull-request (I don't know chinese :).

Thanks, I didn't know this, so will add a remark. Edit: added a remark, I think some examples would help

+1 very simple, very annoying, a real roadblock for pages with multiple inputs. Additionally, because of this label text is misaligned with other elements in other columns and it looks...

+1 to this issue, also concerned that all files should be kept in the memory I think security concerns are easy to overcome by providing hash-based names and remembering mapping...