I have a data set of 10 subjects, each of whom repeated a task five time.
This data set includes two parameters simultaneously Measured. Through an MLP network, I intend to create a kind of regression between two parameters.
Do I have to give the data for each person to the network and average Rmse for all subjects or can all ten subject be entered into a network (10 input) for regression.
A snapshot of your dataset might be helpful to get towards some solution.
I suggest you follow https://www.tutorialspoint.com/managerial_economics/regression_technique.htm
and the attached book for future refence tool.
1. I think amount of data is not sufficient for MLP architecture.
2. It is unclear what you exactly asked. If these two parameters are x and y then I think if you mark the data points for every task you will obtain a chart with 50 points. What you have to do is to fit a specific curve to this chart.
Yes this can be done by MLPs also. Then a single input is related to a specific point in the chart. This means there are 50 such inputs. However you can implement calculations at once using matrix multiplication.
The number of input lines should be same as the number of elements in the feature vector. So for a 10-dimensional FV, there needs to be 10 input lines. The number of outputs should be same as the number of classes. When the first FV is fed to the MLP, the output line of the corresponding class should be 1 while that for the other classes are kept 0 e.g. for three classes the target vector should be [1 ; 0 ; 0]. If each class has five input samples, this is to be repeated 5 times i.e [1, 1, 1, 1, 1 ; 0, 0, 0, 0, 0 ; 0, 0, 0, 0, 0]. Similarly repeat for other classes.
- Badges
- Science topic