Thank you for your reply. Yes, I already know about it but FukuML is not very useful and I have difficulty installing it properly. I was wondering in order to implement the algorithm myself which approach can I use?
You can have a look here: https://github.com/DannyVanpoucke/LSSVMlib I recently implemented an LS-SVM regressor following the API of sk-learn. More info on the actual implementation can be found in this tutorial I created (such that I don't forget why I had to make certain choices.): https://dannyvanpoucke.be/building-scikit-learn-regressor-lssvm-en/
The best implementation was Danny's, but I wanted the speed too. None of the codes can compete with scikit-learn SVM implementation (highly optimized) so I quit looking around.