MACHINE LEARNING OF GRAVITATIONAL WAVES
Using image processing and our CNN algorithm (with 3 convolution layers and 2 hidden layers), we can detect stars based on their color and determine the star's mass from the HR diagram. Therefore, we can predict the maximum and minimum periods of gravitational waves resulting from the rotation of the two stars. The period function will later be plotted against the radius. Apart from the period, we can also estimate the frequency, radiation level, energy, amplitude, and also speed of the star. For further development, we can also detect the distance between two stars so that we can determine the exact period of the two stars. But the current time, human resources, and lack of training datasets are not sufficient to create the project.

TESTING OF BLUE STAR AND YELLOW STAR

We managed to get the precision of the developed algorithm. The level of accuracy and validation accuracy achieved in 20 epochs is 1 for accuracy and 0.92 for accuracy validation