This is my blog.

It’s about machine learning.

Want to learn more? Feel free to ask questions.



What is your blog about?

As I said, it’s about machine learning.

To be more specific, I write about interesting use cases, underrated methods, and ways to easily adapt well-known tools to tackle seemingly not-so-easy problems. In a nutshell, I write about what I wish I could have read 10 years ago, when I was starting out in the field.

Who is it for?

My target audience is anyone with at least a rudimentary understanding of basic machine learning concepts like linear regression, supervised/unsupervised learning etc., and who is interested in learning more. This should hopefully include some of my students, as the blog also serves as additional support material for classes I teach.

Who are you again?

I’m Yannick Le Cacheux. This is my face:

Weird face

I got a PhD in deep learning in 2020 — so practically the Stone Age given how fast the field moves. Since then, I’ve been working as a freelance consultant in machine learning/data science/AI/whatever for various clients, including LVMH, L’Oréal, Axa, Saint-Gobain, Samsung… As a hobby, I also teach graduate classes in machine learning at CentraleSupélec and Université Paris-Saclay.

Do you write about your work for your clients?

Yes, this blog includes (or will soon include) practical case studies inspired by real work for real clients.

But isn’t it super sensitive?

Yes, but mostly no.

There is obviously no private client data used on this blog: all datasets are either publicly available or artificially generated to somehow resemble real data. The same goes for the technical methods described here: they are typically fairly straightforward—although not necessarily obvious—ideas that have been simplified compared to the actual work done for my clients.

So you won’t find any trade secrets here, sorry.

What’s up with this weird blog name?

Hey, watch your mouth! The domain name is my own company’s, Manifold Technology. Turns out, the initials almost perfectly match a great blog name about machine learning, with just a teeny tiny bit of stretching to make it fit.

Other domain names were expensive, so that’s the one I kept.

Can I hire you as a consultant?

It depends on whether I’m available at the moment. Feel free to drop me an email at legends (at) manifold.fr to discuss this.