Accuracy of prediction model ? MLP node

KNIME Analytics Platform
#1

how to explain that with the same multilayer perceptron MLP model and same training set and test set, the accuracy of the prediction changes each execution of the model.

experiment 1 one hidden layer and 5 hidden neud 500 epoch---> accuracy=94.13%

experiment 2 (I kept the same parameters) one hidden layer and 5 hidden neud ,500 epoch---> accuracy=93,97%

Thanks

 

#2

Have you ticked the box "Random Seed" in the Learner node ?

#3

no I have not check this box ""Random Seed", I must to check the box??

#4

Well, if you want reproducible results for each new run, you should tick it.