Solved

How many training sets are sufficient for neural network with 81 neurons in the input layer?

Posted on 2012-03-18
3
346 Views
Last Modified: 2013-11-25
.... and 1 output, approximately or on average?
0
Comment
Question by:Lamiaaa
[X]
Welcome to Experts Exchange

Add your voice to the tech community where 5M+ people just like you are talking about what matters.

  • Help others & share knowledge
  • Earn cash & points
  • Learn & ask questions
  • 2
3 Comments
 
LVL 31

Accepted Solution

by:
GwynforWeb earned 500 total points
ID: 37735002
My experience of neural networks using back propagation is that you need a huge number training inputs. I have found them disappointing which might be a function of me not using them that much, and as such not that skilled at designing the intermediate layers which can effect the ease with which it can be trained. In short I have found them disappointing.

81 inputs is a lot of inputs and the ability to train the network is related to there being an underlying structure to the input for the network to detect, the more pronounced this structure the easier the network can be trained. Some times pre-processing the input to expose more of the structure before hand can improve training and performance.

A neural network is merely a way of designing a function that gives required output for a given input. There are other ways of designing these functions other than neural networks.
0
 

Author Comment

by:Lamiaaa
ID: 37735071
Thanks, if the input is in the form of matrices, what other "functions" can be used?
0
 
LVL 31

Expert Comment

by:GwynforWeb
ID: 37735693
Unfortunately there is no easy deterministic way of training intelligence into a response function for a given input. The other technique that is commonly used is Fuzzy Logic but without knowing the details of your problem I have no idea if it will work for you.

What ever technique you use will require skill on your part in its use. If you where to post some details of your problem I might be able to make better suggestions. It might well be that Neural Networks are the way to go and you have a design or training problem,

 I am assuming you are using back propagation for training if you have a forward -feeding network or do you have a feedback network?
0

Featured Post

Free Tool: Subnet Calculator

The subnet calculator helps you design networks by taking an IP address and network mask and returning information such as network, broadcast address, and host range.

One of a set of tools we're offering as a way of saying thank you for being a part of the community.

Question has a verified solution.

If you are experiencing a similar issue, please ask a related question

Suggested Solutions

Title # Comments Views Activity
Dice Roll Probabilities 3 149
PHP question(s) about order of output 9 75
Work with App store 7 117
Capture and interpret serial port data 10 53
We are taking giant steps in technological advances in the field of wireless telephony. At just 10 years since the advent of smartphones, it is crucial to examine the benefits and disadvantages that have been report to us.
When we purchase storage, we typically are advertised storage of 500GB, 1TB, 2TB and so on. However, when you actually install it into your computer, your 500GB HDD will actually show up as 465GB. Why? It has to do with the way people and computers…
This tutorial will introduce the viewer to VisualVM for the Java platform application. This video explains an example program and covers the Overview, Monitor, and Heap Dump tabs.
The viewer will be introduced to the technique of using vectors in C++. The video will cover how to define a vector, store values in the vector and retrieve data from the values stored in the vector.

740 members asked questions and received personalized solutions in the past 7 days.

Join the community of 500,000 technology professionals and ask your questions.

Join & Ask a Question