It is understood that most times the output (y) can be dependent on the input (x). My question; is it possible to predict an output with 'no given data' from a set of inputs with available data? For instance, if you want to predict the the torque of an engine (output1, with data) using the fueling level and engine speed as inputs, is it also possible to predict the emission levels (output2, without data) using the same set of inputs?
Thank you!
from my knowledge in neural networking, its not possible to predict a output precisely with only the inputs available.
since the network initially uses a set of data to get trained and understand about the process. Only then the prediction of different values tends to give proper accurate results.
so you need minimum number of output values for different combinations before which you can vary the combination to predict the results.
Thanks Samuel for your wonderful contribution! Although I didn't see any possibility as well, this question was raised due to arguments about my training set and have been looking for references to back up this issue. My department is very curious and inquisitive about neural network and am so thankful for sharing your thoughts!
- Badges
- Science topic
- Similar topics
- Social Psychology
- Attitude