We study the optimization of high-power laser pulse temporal shape to obtain the shortest possible pulse
duration by employing a control feedback loop between a pulse shaper and pulse duration measurements instrument.
Performance of five optimization algorithms were explored - the Differential Evolution algorithm, the Nelder-Mead
algorithm, and three hybridization approaches of Differential Evolution (DE), Bayesian optimization (BO) and Nelder-
Mead (NM) algorithms. The hybrid algorithms perform the exploration with the DE or BO algorithm and the
exploitation with the NM algorithm. Algorithms were tested on a semi-physical model of the actual laser based on
linear and nonlinear phase accumulation. We find that the hybridization of the algorithms yields better overall results.