Machine learning and quantum information
prof. UAM dr hab. Karol Bartkiewicz

Date, Time
22.04, 15:00 - 16:00

Location
Link to MSTeams meeting

Classical programming means writing explicit instructions so that a program processes the input data and correctly answers our questions. Machine learning (ML) is a branch of artificial intelligence research that uses implicit programming, where the program does not receive explicit instructions. This method is particularly suitable for problems that are intuitive to humans but difficult to convert to set of machine instructions. Some complex problems resist known ML methods, especially in quantum systems [1,2]. E.g. designing new drug molecules or supervising quantum communication networks, which under certain assumptions should be protected from eavesdropping by the laws of quantum physics. These tasks quickly become unfeasible as the complexity of the problem increases. Solutions to such problems must be sought using quantum computing for ML [1,3]. This is the original motivation to combine ML and quantum physics [1].

However, there are many other reasons to do so. In particular, ML can be used to motivate theoretical and experimental research in quantum information, quantum state engineering, classification and detection.

To illustrate this, I will discuss a few assorted examples of combining ML and quantum information processing.

[1] J. Biamonte, P. Wittek, N. Pancotti, P. Rebentrost, N. Wiebe, S. Lloyd, “Quantum Machine Learning,” Nature 549, 195-202 (2017).
[2] G. Carleo et al., “Machine learning and the physical sciences,”  Rev. Mod. Phys. 91, 045002 (2019).
[3] V. Trávníček, K. Bartkiewicz, A. Černoch, K. Lemr, “”Experimental Measurement of the Hilbert-Schmidt Distance between Two-Qubit States as a Means for Reducing the Complexity of Machine Learning,” Phys. Rev. Lett. 123, 260501 (2019).