Adaptive Bayesian algorithm for achieving a desired magneto-sensitive transition.

Area of Science:

• Quantum Science
• Precision Spectroscopy

Background:

• Bayesian methods update parameters in quantum science.
• Accurate quantum transition frequencies are vital but hard to achieve due to unknown parameter relationships.

Purpose of the Study:

• To develop an efficient adaptive Bayesian algorithm for finding magneto-sensitive transition frequencies.
• To demonstrate the algorithm's effectiveness in real-time tuning of experimental parameters.

Main Methods:

• Utilized an adaptive Bayesian algorithm to search for optimal conditions.
• Experimentally demonstrated using coherent population trapping in laser-cooled ⁸⁷Rb atoms.
• Controlled transition frequency via magnetic field tuned by d.c. voltage.

Main Results:

• The algorithm converged to the desired voltage from random initial values within few iterations (≥10).
• Achieved a response time of approximately 50 seconds for 10 iterations.
• Demonstrated significant advantages over traditional methods, especially for nonlinear frequency-voltage relationships.

Conclusions:

• The proposed adaptive Bayesian algorithm provides a simple and efficient method for determining transition frequencies.
• This technique has broad applicability in precision spectroscopy, including atomic clocks, magnetometers, and nuclear magnetic resonance.