You are here

Home » Neural Network Forums » Neural Networks » Financial Neural Networks

Noob question

aaronpeterson3's picture
published by aaronpeterson3 on Thu, 08/16/2012 - 16:40

I am gong through the markettrain example in c#.

What I am doing follows the example very much the same way, except I was wondering what I would need to do to pass in an "expected" dataset along with the training? These would be values that the other values would be expected to predict.

I am using TemporalMLDataSet, ElmanPattern and LevenbergMarquardtTraining (just like the example).

I have 4 values and one expected outcome.

Hope there is a way to make this happen.

Neural Network Forums: 
Financial Neural Networks
jeffheaton's picture

Submitted by jeffheaton on Fri, 08/17/2012 - 05:55.

If I understand your question, you are asking why you must pass in expected values? This is required for "supervised training", which is where you are providing labels for a series of inputs. You hope that when the neural network sees other data, similar to your training data, it will identify the new data as similar to training data that you knew the outcome for.

The other side of the coin is unsupervised training. This is where you do not provide expected output. In this sort of training, typically clustering, the machine learning method divides the input data into groups.

Theme by Danetsoft and Danang Probo Sayekti inspired by Maksimer