i want search hyper-parameters in convolutional net built keras. using kerasclassifier , gridsearchcv scikit-learn in line intro given here machinelearningmastery . typically scikit-learn optimizes on 'accuracy' , network runs image segmentation optimizing jaccard index. need define own scoring object grid search using make_scorer explaned here make_scorer , here defining scoring strategy . code section below shows implementation, getting error in model.compile(optimizer=optimizer, loss=eval_loss, metrics=(['eval_func']) , not know specify in metrics. default 'accuracy' assume in case 'eval_func' (which works when not doing grid search) or 'score' neither of these works in case. what right syntax? def eval_func(y_true, y_pred): '''evaluation function dice or jaccard, set global var jaccard=true''' if jaccard: return jaccard_index(y_true, y_pred) else: return dice_coef(y_true, y_pred)...
Comments
Post a Comment