Forecast with neural network or...?

paolo_italy

Hi all,

That's my first post, so please forgive me if I'm too newbie...  ;D

I'v started using RapidMiner Studio since a few days, due to an analysis problem I have.

I have to analyse an industrial transformation process, with ca. 30 input attributes ("raw materials") and ca. 20 output attributes ("product features"). All the attributes are numeric types. In particular, I'd like to produce simulated output by giving a test-set of input attributes.

I thought to neural networks, but up to know I'm going nowhere:

1- Is RM Studio suitable for this kind of simulation?
2- What would you suggest as a modelling alternative to neural networks, if RM doesn't support this kind of task with them?

Thank you all,
-Paolo