Quantum Gaussian Mixture Models - Week 10

This week, I've researched the performance of the five clusters case. I generated the data set using mlpack GMM class. When training, the value 0 was good as an initial difference of phi. From some experiments, I figured out that as the difference neared zero, the two clusters showed a tendency to move away, while as the difference neared negative one, the two clusters tended to be close. Lastly, I think it would be nice to take a closer look at the correlation between phi and other parameters to utilize the variation of phi more efficiently.

Thanks for reading :)