%info@http:://golem.fjfi.cvut.cz/wiki/root/GW4reports
\def\GWslide{\slide{Bayesian Discharge Optimization}{
\begin{columns}[c]
\column{0.5\tw}
   \GWif{width=\tw}{/Experiments/TokamakRegimes/BayesianOptimization/Ficker-2025-FYS1/BayesianOptim.png}{Domain for the optimization of the Golem discharge length}
\column{0.5\tw}
\begin{itemize}
  \item Artificial intelligence (Bayesian optimization) is used to automatically improve tokamak discharge settings.
  \item The algorithm learns from previous discharges and decides which parameters to try next.
  \item The first objective was to maximize the discharge duration using basic GOLEM control parameters.
  \end{itemize}
\end{columns}
\begin{itemize}
  \item AI-driven optimization led to a stable and reproducible long-discharge regime (up to 26.3\,ms).
  \item This approach demonstrates how AI can support more advanced plasma control in future experiments.
\end{itemize}
}}