Michael's and Christian's Blog

Mindblown: a blog about philosophy.

  • Interactive Plotting Backend for Model-Diagnostics

    With its newest release 1.1.0, the Python package model-diagnostics got the concept of a plotting backend. Before this release, all plots were constructed with matplotlib. This is still the default. But additionally, the user can now select plotly, if it is installed. There are 2 ways to specify the plotting backend explicitly The context manager…

  • Building Strong GLMs in Python via ML + XAI

    We use Python to craft a strong GLM by insights from a boosted trees model.

  • ML + XAI -> Strong GLM

    In this post, we improve a simple GLM by insights from a boosted trees model.

  • Explain that tidymodels blackbox!

    In this post you will learn how to explain a {tidymodels} blackbox with classic XAI and SHAP.

  • An Open Source Journey with Scikit-Learn

    In this post, I’d like to tell the story of my journey into the open source world of Python with a focus on scikit-learn. My hope is that it encourages others to start or to keep contributing and have endurance for bigger picture changes.

  • Permutation SHAP versus Kernel SHAP

    When do the two methods agree? When not?

  • Interactions – where are you?

    This question sends shivers down the poor modelers spine… The {hstats} R package introduced in our last post measures their strength using Friedman’s H-statistics, a collection of statistics based on partial dependence functions. On Github, the preview version of {hstats} 1.0.0 out – I will try to bring it to CRAN in about one week…

  • It’s the interactions

    What makes a ML model a black-box? It is the interactions. Without any interactions, the ML model is additive and can be exactly described. Studying interaction effects of ML models is challenging. The main XAI approaches are: This post is mainly about the third approach. Its beauty is that we get information about all interactions.…

  • Model Diagnostics in Python

    Version 1.0.0 of the new Python package for model-diagnostics was just released on PyPI.

  • Geographic SHAP

    “R Python” continued… Geographic SHAP

  • Quantiles And Their Estimation

    Applied statistics is dominated by the ubiquitous mean. For a change, this post is dedicated to quantiles. I will give my best to provide a good mix of theory and practical examples. While the mean describes only the central tendency of a distribution or random sample, quantiles are able to describe the whole distribution. They…

  • SHAP + XGBoost + Tidymodels = LOVE

    tidymodels and shapviz to explain XGBoost models

Got any book recommendations?

Get In Touch