How can I set targets in nueral network pattern recognition?

Hi friends,

I am a new Matlab user. Can you please help me to guide for targets in neural network pattern recognition? How can I set targets for pattern recognition? For example, I have some sample data of machine monitoring like; temperature, vibration, and pressure of machine monitoring data. How can I use this data in classification and to set targets for it? I have per minute monitoring data, as; at 1min T= 65c , V= 40mm/sec , P = 50psi (data is in good range means OK (1) ) and at 2min T= 90c , V= 40mm/sec , P = 48psi (temperature is not in OK condition but others are OK) At second point, should I consider it 0 or 1 for targets? Thanks

Answer 1

[Greg Heath]

If your {0,1} answer only depends on whether one or more of three inputs is out of range, then there is no need for a NN. On the other hand, a NN is reasonable if there are certain regions of 3-D space corresponding to target 0.

[Umair Sarwar]

Thanks Greg heath for your prompt answer. Will you please guide me further, which algorithm I should use for classification? Thanks