Black holes in orbit around each other emit gravitational waves. This causes the orbit to become faster and tighter, and eventually, the black holes merge in a final burst of radiation. These gravitational waves propagate through the Universe at the speed of light, and are detected by observatories in the USA and Italy.
Scientists compare the data collected by the observatories against theoretical predictions to estimate the properties of the source, including how large the black holes are and how fast they are spinning. Currently, this procedure takes at least hours, often months.
An interdisciplinary team of researchers in Potsdam is using machine learning methods to speed up this process. They developed an algorithm using a deep neural network, a complex computer code built from a sequence of simpler operations, inspired by the human brain. Within seconds, the system infers all properties of the binary black-hole source.
This journal cover illustration’s goal was to show the interaction between gravitational wave patterns and the machine learning algorithm.
On the left, the gravitational waves produced by a merger of two balck holes is depicted with a mysterious purple color. This shows that black holes are still mysterious objects for our understanding, and that gravitational wave patterns can be difficult to interpret.
On the right, the human brain shape represents the deep neural network used for the analysis. The dots and lines represent the simpler processes that make up the network and that interact in a more complex way.
Finally, the two elements are connected with longer lines, which show the neural network’s capability to extrapolate information efficiently. The brighter color of the neural network symbolizes how this technique can allow us to shed light on our universe.