Monthly Archives: June 2025

Fellow machine learning enthusiast Thorsten Kleppe (https://github.com/grensen, thorsten.kleppe@gmail.com) produces amazingly beautiful data and machine learning visualizations. He sent me an email message recently with some of his latest explorations: Stunningly beautiful. And the visualizations are interactive — the viewer can … Continue reading →

I wrote an article titled “Linear Support Vector Regression Using C# with Particle Swarm Training” in the June 2025 edition of Microsoft Visual Studio Magazine. See https://visualstudiomagazine.com/articles/2025/06/03/linear-support-vector-regression-using-c-with-particle-swarm-training.aspx. There are approximately a dozen common regression (predict a single numeric value) techniques. … Continue reading →

I spend some time every morning before work to learn something new. For me, learning something new usually means writing a computer program of some sort. On a recent morning, I figured I’d demonstrate how to compute the roots of … Continue reading →

Kernel ridge regression (KRR) is a technique to predict a single numeric value. KRR uses a kernel function, which compares two vectors and computes a measure of their similarity in order to handle complex non-linear data, and the ridge technique … Continue reading →

Kernelized support vector regression is a rarely used technique to predict a single number. One of the problems with kernelized SVR is that it uses a loss function (epsilon-insensitive loss) that is not calculus-differentiable and so ordinary training/optimization techniques such … Continue reading →