optbinning

optbinning copied to clipboard

KS using .predict_proba() and .score() of Scorecard is different

3

Hi @guillermo-navas-palencia ! As always, congratulations for the amazing job on the library! I was doing some tests using the Scorecard class and noticed that de KS statistics reported when...

ds-mauri

Use bins thresholds values to labels the bins

1

Hi, currently there is two options for plots: normal where bins are labeled with bins numbers and 'actual' where the real scale is used. I'd like something in between where...

lcrmorin

Similar to: http://gnpalencia.org/optbinning/binning_scenarios.html

guillermo-navas-palencia

Add the following constraints to all OptBinning classes: - ``min_n_neg_splits``: minimum number of negative split points. - ``max_n_neg_splits``: maximum number of negative split points. - ``min_n_pos_splits``: minimum number of positive...

guillermo-navas-palencia

guillermo-navas-palencia

guillermo-navas-palencia

Model variable screening

2

How to develop a scorecard that uses lasso or ridge for variable screening to get a model that is more generalizable than a model with a full subset of variables？

cfkstat

guillermo-navas-palencia

Develop a heuristic method to obtain good feasible solutions faster. - Greedy approach with variable neighbourhood search methods. - Prioritize features with higher importance. - Optionally, use the result as...

guillermo-navas-palencia

Add support to HIGHS MIP solver via scipy.optimize.linprog API. This should be able to generate models significantly faster than ortools. Similar performance to CP-SAT solver is expected for small and...

guillermo-navas-palencia