RCC, MIPT, MISIS and Skoltech: An algorithm capable of predicting the behavior of quantum systems

The main articles are:

• Big Data
• Data mining Data mining
• Predictive analytics (predictive, predictive, predictive) Predictive analytics
• Quantum computers and networks in Russia
• Quantum computer and quantum communication

2022: Application of predictive analytics in quantum technologies

On November 1, 2022, it became known that a group of scientists from the Russian Quantum Center, the Moscow Institute of Physics and Technology, MISIS University and the Skolkovo Institute of Science and Technology, together with colleagues, developed an algorithm capable of predicting the behavior of quantum systems subject to their interaction with the external environment. The predictive approach is entirely based on the analysis of available data, which is comparable to predictive analytics based on "big data," often used by technology companies. The results of numerical experiments are described in the scientific journal Physical Review Research.

Quantum computers are computing devices capable of solving certain classes of problems orders of magnitude faster. This feature is achieved by using two quantum effects: superposition and quantum entanglement. Thus, unlike classical bits receiving state 0 or 1, quantum bits (qubits) can be in all states at once. However, qubits are subject to external factors, including noise, which leads to errors in calculations.

A group led by Alexei Fedorov has developed an algorithm that, according to the observed behavior of a quantum system, allows predicting its further dynamics. Of particular importance to the scientific community is the fact that the algorithm makes it possible to predict the behavior of quantum systems with memory - one of the most complex effects, also known as "non-Markov behavior." Predicting the behavior of such non-Markov systems is extremely difficult, since it is impossible to measure the parameters of their environment or accurately restore the nature of their interaction with the environment. However, such a task is key for building quantum computers.

It is expected that the developed method will be highly in demand in the field of both quantum computing and quantum sensors. It can be used by researchers to accurately diagnose the functioning of individual elements of quantum devices: qubits - in quantum computers, sensitive elements - in quantum sensors. The results were obtained with the support of the RPF and other programs.

The study applied methods actively used in solving classical data processing problems to build a model of a complex quantum system. The algorithm allows you to "learn" the behavior of a quantum system interacting with the environment and predict its further dynamics over time. The developed approach captures all the effects of the interaction between the system and the environment, while the method does not require direct access to the environment - enough data on the dynamics of the system, from which information is extracted and about the environment. In the future, the method can be used to solve problems of quantum technologies, noted Ilya Luchnikov, researcher at the Quantum Information Technologies group of the Russian Quantum Center and MISIS University.