INFORMATION TECHNOLOGY

Scientists demonstrate the quantum advantages of information storage for stochastic modeling of complex systems


Professors Li Chuanfeng and Xiang Guoyong and collaborators in the team of academician Guo Guangcan of the University of Science and Technology of China can achieve higher accuracy than any classical model of the same memory dimension using the memory of a single qubit. The study demonstrates the storage advantages of quantum technology in the modeling of non-Markov processes of complex systems. On May 6, the research results were published online in Nature Communications.

Conceptual diagram of classical clock process simulated by quantum model in experiment Courtesy of China University of Science and Technology

From chemical reactions to financial markets, from meteorological systems to galaxy formation, people need to deal with complex processes of all scales. Stochastic modeling can help us predict the future behavior of these processes. However, since these stochastic processes are generally non-Markov, their future behavior depends not only on the current state, but also on its past state. In order to simulate such a process, there must be a memory to store a large amount of observational information in the system. The amount of information stored will be directly related to the accuracy with which future behavior is predicted, so this will lead to a bottleneck in practice that requires a trade-off between memory reduction and prediction accuracy.

Xiang’s research group found that quantum technology can demonstrate significant advantages even when modeling purely classical dynamic processes. The team implemented a series of quantum models of non-Markov stochastic processes based on photonic system experiments. This class of random procedures has a tunable parameter that controls its effective memory length, and the memory dimension of the optimal classical model grows with the value of this parameter. Experiments have shown that a quantum model can simulate any process in this behavior (i.e., any of this parameter) using only a single qubit as memory, and even in the presence of experimental noise, this quantum model can make predictions of future behavior more accurately than the optimal classical model of the same memory dimension.

The editors of Nature Communications commented on the work: “Quantum technology has a storage advantage in simulating stochastic processes, but experimental verification in non-Markov processes (where the quantum advantage is stronger) has not been achieved. The authors of this paper use single-bit storage to model non-Markov processes, filling this gap through theoretical analysis and experimental verification. ”

According to the researchers, the work is a key step towards demonstrating the scalability and robustness of the advantages of this quantum memory. (Source: Wang Min, China Science News)

Related paper information:https://doi.org/10.1038/s41467-023-37555-0



Source link

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button