PinnedKrzysztof JoachimiakInstance importance for regression with AXIL weightsIntroducing a novel class of XAI methods.8 min read·Jul 1, 2023----
Krzysztof JoachimiakLearning to play Pong with PyTorch & TianshouTraining a Deep Q Network with a modular RL library 12 min read·Oct 13, 2023----
Krzysztof JoachimiakThe forgotten initial estimator in GBMsDo you know, how to extract the initial guess from popular GBM models?6 min read·Jun 24, 2023--1--1
Krzysztof JoachimiakDiscretizing features with gradient boosted decision treesDistilling discretization splits from GBM models7 min read·Jun 17, 2023----
Krzysztof JoachimiakExtracting trees from GBM models as data framesscikit-gbm provides a unified API to extract trees from the most popular GBM models4 min read·Jun 11, 2023--1--1
Krzysztof JoachimiakFeaturizing text for regression — Part I: scikit-learnClassical and easy-to-use methods to handle text features in regression tasks9 min read·Jun 7, 2023----
Krzysztof JoachimiakForecast time series like a ChatGPT Prompt EngineerWhat if we ask ChatGPT to guess future values of a time series?6 min read·May 6, 2023----
Krzysztof JoachimiakGenerating features with gradient boosted decision treesI’m not the first person, who publishes an article on that topic on Medium. There is already at least one similar article by Carlos Mougan…4 min read·Apr 19, 2023--2--2
Krzysztof JoachimiakMulti-output regression with gradient boosting machinesThe first type of models that comes into my mind when thinking about the multi-output regression is definitely not the GBM. I’d rather…16 min read·Apr 14, 2023--4--4
Krzysztof JoachimiakExpectile regression: an alternative for quantile regression?The quantile regression is not the only way to estimate the prediction lower and upper bounds.8 min read·Mar 27, 2023----