Technologies that take advantage of new behaviors in quantum mechanics are likely to become mainstream in the near future. These can include devices that use quantum information as input and output data, which require careful verification due to inherent uncertainties. Verification is more difficult if the device is time dependent when the output is dependent on past inputs. For the first time, researchers using machine learning have dramatically improved the efficiency of verification of time-dependent quantum devices by incorporating some memory effect present in these systems.
Quantum computers are making the headlines in the scientific press, but these machines are considered by most experts to be still in their infancy. A quantum internet, however, may be a little closer to the present. Among other things, this would offer significant security advantages over our current Internet. But even that will rely on technologies that have yet to see the light of day outside the lab. While many fundamental building blocks of the devices that can create our Quantum Internet may have been developed, there are many technical challenges in realizing them as products. But a lot of research is underway to create tools for the design of quantum devices.
Postdoctoral researcher Quoc Hoan Tran and Associate Professor Kohei Nakajima from the Graduate School of Information Science and Technology at the University of Tokyo pioneered such a tool, which they believe could make behavioral verification of quantum devices a more efficient and precise company than it is. is at present. Their contribution is an algorithm that can reconstruct the functioning of a time-dependent quantum device by simply learning the relationship between quantum inputs and outputs. This approach is actually common when exploring a classical physical system, but quantum information is usually tricky to store, making it usually impossible.
âThe technique for describing a quantum system based on its inputs and outputs is called quantum process tomography,â Tran said. “However, many researchers now report that their quantum systems exhibit some kind of memory effect where current states are affected by previous states. This means that a simple inspection of the input and output states cannot describe the time dependent nature of the system. You could model the system multiple times after each time change, but this would be extremely computationally inefficient. Our goal was to embrace this memory effect and use it to our advantage rather than to use brute force to overcome it.
Tran and Nakajima turned to machine learning and a technique called quantum reservoir computation to create their new algorithm. It learns patterns of inputs and outputs that change over time in a quantum system and effectively guesses how those patterns will change, even in situations the algorithm has not yet experienced. Since it does not need to know the inner workings of a quantum system like a more empirical method would, but only the inputs and outputs, the team’s algorithm can be simpler and produce results faster. .
âRight now our algorithm can emulate some type of quantum system, but hypothetical devices can vary greatly in their processing capacity and have different memory effects. So the next step in the research will be to expand the capabilities of our algorithms, essentially doing something more general and therefore more useful, âTran said. âI’m excited about what quantum machine learning methods could do, what hypothetical devices they could lead to. “
Reference: “Learning Temporal Quantum Tomography” December 22, 2021, Physical examination letters.
DOI: 10.1103 / PhysRevLett.127.260401