Hello all,
I want to create a model in visual basic taking data from Access or a project in Excel using VBA both are fine. The model should be able to predict confidence intervals on the test data of the neural network by the technique explained in "G. Chryssolouris, M. Lee, and A. Ramsey, *Confidence interval prediction for neural network models, IEEE Transactions on Neural Networks*, v 7, n 1, Jan, 1996, pp 229-232."
Of course, I cannot give you the paper but I've documentation of my own and I'm attaching it. If you know NN and if you follow the documentation it's easy to develop the model. If you have any questions please do ask me.
The documentation looks difficult but it's not, there is no need to calculate the derivatives in that, the neural network itself calculates the derivatives.
The neural net architecture is one input layer -one hidden layer-one output layer with one output neuron
The requirements of the project are:
Program should be able to take input from the user in a window form.
The inputs to the program are learning rate, number of hidden neurons, number of input neurons, total number of observations, percentage to be trained/tested, number of iterations, preferably it should ask for the data file and the table name, it it's coded in Excel/VBA then Sheet name and data range.
There is some matrix manipulation involved and for that one module is present in planet source code
Also, before training the data should be normalized and the output should be again un-normalized.
I've already developed the model in VB6 but it's not giving me the results I want. So I want someone else to do it then I can compare my code. I can provide you bits of my code if you'd like.
## Deliverables
1) Complete and fully-functional working program(s) in executable form as well as complete source code of all work done.
2) Deliverables must be in ready-to-run condition
3) All deliverables will be considered "work made for hire" under U.S. Copyright law. Buyer will receive exclusive and complete copyrights to all work purchased. (No GPL, GNU, 3rd party components, etc. unless all copyright ramifications are explained AND AGREED TO by the buyer on the site per the coder's Seller Legal Agreement).
4) The program should be user friendly.
5) The plot of original data, predicted data and the confidence intervals should be shown at the end of the program.
6) Preferably training curve should be shown.
## Platform
Windows XP, 2K, 98.