About MachineCurve

Me after getting my master’s diploma.

Hi there, and welcome to MachineCurve!

My name is Chris. Actually, it’s Christian, but in my opinion Chris is better for the sake of pronunciation. I live in the Netherlands and have a background in business informatics, which means that I am formally trained to understand what happens in organizations and how IT can be applied to make business better.

However, if I have to be honest, I’m much more technical than the average business informatics students at my alma mater. Having started programming in 2007, I’ve learnt over the years how to create websites, how to create scalable web applications, how to dive deep into problems with solutions far out of sight and how to set up software architectures and cloud infrastructure to support these. I’ve had one great mentor for this over the years, which I’m really thankful for.

In 2014, during my bachelor programme at university, I was introduced to Artificial Intelligence. For a study assignment, I had to build a Naive Bayes Classifier from the ground up. And fun it was! Ever since, I’ve been fascinated about how machine learning can seemingly learn things humans can do too. Unfortunately, picking up machine learning was difficult back then: I had study related work to do, worked as a freelance software engineer, thought it would be important to be a good friend and partner to my girlfriend, and so on. Since machine learning theory is not the easiest matter for digestion, I didn’t really work on machine learning after the study project.

Until April 2018. During that month, I was asked to participate in a project that researched opportunities for discriminating material type from Ground Penetrating Radar (GPR) signals, which are used for detection of objects in the underground, e.g. for avoiding utility strikes. The challenge was enormous, but we started anyway. In four weeks, we built a model that worked but – with today’s new insights – was a really poor try in terms of architecture (TLDR: SVMs with linear kernels for highly nonlinear data … oops ๐Ÿ™Š).

But after the project, I finally found the time to pick up machine learning and deep learning theory. Especially during Jan – Aug 2019, when I worked on my master’s thesis that extended the GPR project (this time with good results both in terms of ML outcomes and appreciation by the professorate), I’ve learnt a lot. That’s why I picked up working on MachineCurve in May 2019 (which already existed since September 2017, but only with two blogs on it). The goal: to make available my learnings at global scale. And likely, that’s why you’re reading this introduction.

So, a long story made short: welcome to MachineCurve!

It’s a website where you’ll find my learnings made available for your own consumption. Whatever I find interesting, is what I will cover in various blogs on machine learning. They range from traditional ML techniques such as Rosenblatt Perceptrons or SVMs to newer deep learning techniques, such as Variational Autoencoders and GANs.

If you have any questions or if you wish certain content to appear on my website, feel free to contact me at chris followed by at and the domain name of this website. You may also add me on LinkedIn (please include a message why you’d like to connect, to avoid me thinking it’s a spammy account) if you think that would benefit you, me or the both of us. And if you’d like to stay up to date about content on this website, feel free to subscribe to the MachineCurve Community newsletter through the popup on the right. Joining is free and unsubscribing can be done at any time, but I’ll do my best to avoid that being necessary ๐Ÿ˜„

All right. Thanks for reading MachineCurve today, happy browsing and… happy engineering! ๐Ÿ˜Ž