Introduction to Neural Networks
Author: Mockasino
Creation Date: 12/25/2016 7:15 PM
profile picture

Mockasino

#1
Hello,

i am new to Neuronal Networks.

Has anybody here in the Forum, who is using the Neuronal Network and the WL Neuro-Lab for developing trading strategies, a hint regarding literature for NN and trading strategies?
What will generally be a good "start" on this Topic NN?
Any recommendation?

Thanks in advance!

Regards
Christian
profile picture

Eugene

#2
Christian,

Here is a good place to get started:

List of Bruce Vanstone's publications (PDF) on neural networks

Choose "Neural networks (9)" to filter the list. Bruce was used to be an avid NN popularizer. Next, check out the Neuro-Lab Guide (NN > Help > Contents) and follow posts by LenMoz like this one who is the NN expert.
profile picture

Mockasino

#3
Thanks Eugene,

These are good sources .

One question to the NN-Indicator in WL. How is it calculated?
And how do i have to deal with NN-Indicator in strategies in combination with number of Observation.

Maybe these are stupid questions, but i did not catched this really in WL-help.

Regards
Christian
profile picture

Eugene

#4
QUOTE:
One question to the NN-Indicator in WL. How is it calculated?

Quoting the NN User Guide:

QUOTE:
NN Indicator

This column contains the values that the neural network produced. The range of values corresponds with the observations in the same row. Neuro-Lab indicators always produce output in the range of 0 to 100. When a network outputs a low value it is predicting that the output will be close to low end of the output range. Likewise, when the network outputs a high value, it is predicting that (based on the given inputs) the output will be in the high end of the range. For example, if the Output Script is set to predict the net return after 5 days, then the NN Indicator value will be higher when the NN thinks that there will be a higher profit after 5 days.
This website uses cookies to improve your experience. We'll assume you're ok with that, but you can opt-out if you wish (Read more).