how to create an own scoring with the output of decision_function(X_val)?

[fcoppey]

I would like to use LocalOutlierFactor.decision_function(X_val) in my optimization. is it possible?

Do I have to create an own scoring function for that? If yes, any idea how? I've been looking for some time and can't seem to find a solution.

here is my code :

param_grid = {'n_neighbors': range(2, min(X_train.shape[0], 30)), 'contamination': np.linspace(0.01, 0.2, 15), 'novelty': [True]}
gs = GridSearch(model=LocalOutlierFactor(), param_grid=param_grid)
gs.fit(X_train_, y_train, X_val, y_val, scoring=??what goes here??)

Thanks for your help