Studied Mathematics, graduated in Cryptanalysis, working as a Data Scientist. Interested in algorithms, probability theory, and machine learning. Python user.

A short Story

The three friends Frequentist Frank, Stubborn Stu, and Bayesian Betty go to a funfair where a mysterious-looking tent catches their eyes. Inside, they meet Claire Voyant who claims to be a… fortune teller. The friends don’t believe her, of course — they need proof. So they conduct a little experiment:

Bayesian Marketing Mix Modeling in Python via PyMC3

Estimate the saturation, carryover, and other parameters all at once, including their uncertainty

In this article, I want to combine two concepts that I discussed in earlier posts: Bayesian modeling and marketing mix modeling. Since the chances are high that you are not familiar with both of these topics, let me give you a quick introduction and further readings. I will

1. motivate what…

Interpretable Neural Networks With PyTorch

Learn how to build feedforward neural networks that are interpretable by design using PyTorch

There are several approaches to rate machine learning models, two of them being accuracy and interpretability. A model with high accuracy is what we usually call a good model, it learned the relationship between the inputs X and outputs y well.

If a model has high interpretability or explainability, we…

Rockin‘ Rolling Regression in Python via PyMC3

Learn how to deal with varying parameters

Assume that you want to train a parametric model such as a linear one or a neural network. In the case of linear regression, first, you specify the shape of the model, let us say y = ax + b. Second, you estimate the parameters a and b. …

Flashback

In my last article, I introduced you to the world of marketing mix modeling. If you have not read it so far, please do before you proceed.

There, we have a created a linear regression model that is able to predict sales based on raw advertising spends in several advertising…

Introduction to Marketing Mix Modeling in Python

To keep a business running, spending money on advertising is crucial — this is the case regardless of whether the company is small or already established. And the number of ad spendings in the industry are enormous:

A Tale of Two Approaches

From time to time, you have the agony of choice when trying to do machine learning on a quite heterogeneous dataset. As an example of what I mean with a heterogeneous dataset, consider a simple height of people dataset.

Staying Competitive with Linear Models

With the right features, a linear model can become a beast

As a data scientist, you have a lot of different types of shiny machine learning models to choose from. There are neural network models, gradient boosting models, bagging models, just to name a few main categories.

In the shadows of these fellas, there also exist models like linear and logistic…

Introduction

Machine learning is often all about the following question:

Given a dataset (X, y), where X is a feature matrix and y is a target vector, find an f with f(X) ≈ y.

We usually do not enforce a strict equality à la f(X) = y because there are errors…

Introduction to Linear Programming

Recognize linear program problems and solve them in Python with CVXPY 