Saving / restoring a group of windows & Neural network question

noobie questions-

  1. Does Amibroker support opening and closing a group of windows
    my use case-

    I want to use 3 windows with a unique set of indicators when evaluating
    portfolio -1
    and a separate set of windows and indicators when evaluating portfolio-2.

  2. Does amibroker support creation of neural nets?

  • Grassopah!

@lilgrasshopah you can save your preferred chart setup as a layout, worth carefully learning from the User Guide

As for your second question, AmiBroker’s afl language enables you to code neural nets if you know how to write the code. A couple of simple examples,

Hope you share your NN expertise and code.


Thank you @portfoliobuilder for that link which provided the answer.
Part of my challenge is translating terminology from the platform I am coming from to AmiBroker terminology.

Re: NN - As I am evaluating Amibroker I do not have access to the members area.
My question on NN should have been:

  • Does Amibroker version 6.30+ ( I am using ver 6.0) offer an NN creation wizard?

There is no "wizard" for neural networks.

@lilgrasshopah, as stated by @Tomasz, there is no native wizard, but the first AFL example linked by @portfoliobuilder was generated using a 3rd party plugin + an "addon"1) that has a wizard to create NN code for AmiBroker.

No recommendation here, since I do not know nor the product nor the developer, but the available documentation provided online for the plugin, and the NN wizard should give you a pretty good idea of the capabilities.

If it is the kind of thing you are looking for, you should directly contact the plugin developer to get more information and to check about the compatibility with the current versions of AmiBroker.

1) This plugin is developed/sold from an INDEPENDENT business that has no financial/engineering or other connection to


Hello @lilgrasshopah first you are welcome,

Regarding NN I think you have to train your network on data carefully first in more specific program and then once you have good prediction ability AFL is dynamic and strong enough to program such network since all the rest is about looping and matrices manipulation and AFL contains plenty of the functions that can deal with that.

BTW, take care of relying on prediction of NNs in stock market unless you train it very well and the deviation of results is insignificant !!!!!