Regression Learner App Vs. Neural Fitting App

Hi,

I have used both "Regression Learner App" and "Neural Fitting App" for my 14 variables - 1 target regression model. It seems the models' performance somehow similar. So is there a big difference between the two? Besides the fact that NFA uses the neural network and its algorithms. (But RL app is machine learning).

And when I use both models to predict a new dataset for unknown results, the model of Neural Fitting generate negative values which are not right for my dataset. It is interesting.