Expectation Maximization C++ code review

As a side project to get into machine learning I am trying to implement an expectation maximization algorithm in C++ but I am going wrong somewhere.

Would anyone be willing to look over my code and help me understand where I am going wrong?