py_default_default_5.stdout

Hyperparameter optimization for MLP:
Best cross-validation score: 0.7160
Best parameters:
{\small
\begin{verbatim}
{'alpha': 0.01, 'hidden_layer_sizes': (128, 64, 32), 'learning_rate_init': 0.01}
\end{verbatim}
}

Top 5 parameter combinations:
{\small
\begin{verbatim}
alpha=0.01, layers=(128, 64, 32), lr=0.01 : 0.7160 $\pm$ 0.0152
\end{verbatim}
}
{\small
\begin{verbatim}
alpha=0.0001, layers=(64,), lr=0.01 : 0.7140 $\pm$ 0.0160
\end{verbatim}
}
{\small
\begin{verbatim}
alpha=0.001, layers=(64,), lr=0.01 : 0.7140 $\pm$ 0.0160
\end{verbatim}
}
{\small
\begin{verbatim}
alpha=0.01, layers=(64,), lr=0.01 : 0.7140 $\pm$ 0.0160
\end{verbatim}
}
{\small
\begin{verbatim}
alpha=0.0001, layers=(128, 64, 32), lr=0.001 : 0.7081 $\pm$ 0.0361
\end{verbatim}
}

Random Forest feature importance analysis:
  Top feature importance: 0.0308
  Mean feature importance: 0.0156
  Features with zero importance: 0