Discussion of article "Practical application of neural networks in trading (Part 2). Computer vision" - page 3
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Published article Practical application of neural networks in trading (Part 2). Computer vision:
Author: Andrey Dibrov
The article Practical application of neural networks in trading (Part 2) has been published . Computer Vision:
Author: Andrey Dibrov
You'll forgive me, but in the presence of digital data to process graphics - it's a perversion...
1. You should do everything in png =)
2. For the purposes of neurotrending, reinforcement learning is suitable.... Otherwise you will have to explain to neurons the difference between two graphical images (arrays), and in the presence of digital data this is an unnecessary trick. Neurons understand everything digitally and source data digitally too =)
I'm sorry, but with digital data, it's perverse to process graphics.
Forgive me, but with digital data, processing graphics is a perversion....
1. You have to do everything in png =)
2. For the purposes of neurotrending, reinforcement learning is suitable.... Otherwise you will have to explain to neurons the difference between two graphical images (arrays), and in the presence of digital data this is an unnecessary trick. Neurons understand everything digitally and source data digitally too =)
Well.... I use graphics to identify similarities and then use them as a filter. That's all. I would advise you to also use neurons to analyse text messages...
The point is not to formalise something into "chisels". Zigzag is a problematic indicator in general... Specifically lagging in dynamics and not telling anything....
https://youtu.be/mcQH-OqC0Bs, https://youtu.be/XL5n4X0Jdd8
you're dead wrong on this one. Stop pipsing with the review of one candle and everything will be immediately obvious.
you're dead wrong. Stop pipsing with the review of one candle and everything will be obvious at once.
If we look with the best neural network in our head at the screenshot of the chart, which our artificial neural network is reviewing.... we can see that there's more than one candle on it.
I'm sorry, but with digital data, it's perverse to process graphics.
Forgive me, but with digital data, processing graphics is a perversion....
1. You have to do everything in png =)
2. For the purposes of neurotrending, reinforcement learning is suitable.... Otherwise you will have to explain to neurons the difference between two graphical images (arrays), and in the presence of digital data this is an unnecessary trick. Neurons understand everything digitally and source data digitally too =)
And yes, at the training stage, the hardware resources need to be raised significantly. But this is all within reason.
But at the stage of analysis and response, a neural network needs only an image without additional digitised data. For example, my working neural networks, which analyse time series, are lined up in a chain and each has more than 50 inputs.
So here is the question - where is it better to twist...? At the training stage or at the working stage.