The prediction parameters training tab is used to set up parameters for the neural network training. For more information on training parameters see Neural Network – Network Criteria Discussion (Advanced).

The parameters on this tab may or may not be visible depending upon the currently selected Prediction Wizard Interface Options. Press the Options button to change the Prediction Wizard Interface Options. For more information see Prediction Wizard Interface Options.

1. Select the Training objective used to select optimal network structure and optimal inputs/parameters.

This is the objective that the prediction uses to determine the “best” prediction during training and optimization. For more information see Trading Objective Functions and Error Objective Functions

If you select the minimize error objective or another non-trading objective, trading statistics for the net can still be calculated based on trading rules. You can still choose whether you want trading simulated with long positions, short positions, or both on the Positions tab.

2. Setup the Advanced training parameters.

Number of hidden nodes during training – This will determine how many hidden nodes to use when training and optimizing.

Adjust training set for trending markets by evenly distributing training bars ‘ This will make the neural net train on an equal amount of positively trending data points and negatively trending data points. Warning: This randomly removes training cases that trend in the more populated direction, and thus the neural network will have fewer cases from which to learn. The training will occur on fewer points, but the statistics will be calculated over all of the data points.

When you are satisfied with the prediction parameters to be used when training and optimizing press the OK button to return to the Prediction Wizard.

Notes:

• Not all objectives may result in the type of trading you like to accomplish, and how they perform could depend upon many factors, like the issue you are trading, the inputs you have chosen, the time periods chosen, and many other things. If some objective functions don’t perform for you, try some others.

• During optimization the objective function is calculated at exactly the Number of hidden nodes during training. During training the objective function is calculated for 0 hidden nodes to the Number of hidden nodes during training, and the number of hidden nodes with the best results is used. This means that even though the number of hidden nodes during optimization is exactly the same as the number of hidden nodes during training, the training results may be better than the optimization results on the same training set.

Topics of Interest:
What are Neural Networks?
Neural Network – Network Criteria Discussion (General)
Neural Network – Network Criteria Discussion (Objective)
Neural Network – Network Criteria Discussion (Advanced)
Troubleshooting Your Model – What to Do if You Feel You Haven’t Been Successful
Using Predictions