This is my first post here in this section. I have built a C-library for computing matrices and I going to use that for system identification, which IMO i think is a part of machine learning, even if system identification has developed separately from machine learning.
The reason why I selected C language is not because I think it's the best language and C++ is worst language. No, I think C is a bad language compared to other languages such as C++, Python and Java, and those are much better than C. But when it comes to reality and jobs, there is no much options than C, MATLAB, Java, C++ and C# for the industry. Java is my favorite language due to its power to create webapplications. Closed software is the key for the bussnies in my case.
So anyway! I developed a C-library and it works fine to compute matrix algebra. I will apply this C-library for embedded systems, which was the goal of creating this library.
When I want to learn machine learning and deep learning, I will focus that on control theory. My goal with machine/deep learning is to learn machines to make a decision by analysing data.
"Where should I turn left" "What lamp should I light" "When should I decrease or increase" "What of message should I send"
I'm not into pattern recognition of numbers and picture analysis. That is to much for me. I want to use the small methods that requries small amout of data, e.g 150 data vector or 20x20 matrix.
So that's leads me to my opening question: Is there any math-methods I can apply for machine/deep learning with my linear algebra library, written in C. Where that method works for embedded system with the goal of turn control theory into practise?
My library is here: https://github.com/DanielMartensson/EmbeddedAlgebra