NS + indicators. Experiment. - page 10
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
What's the second way?
Right-click on the SOC node, select "details", "object" tab, "as text" button - and you see all the same information as in the .ded file, but without the tags. You can select the text and copy it to the clipboard . Exporting this information is disabled in the academic version.
Just to complete the picture. And also information for reflection.
In one of Larry Pesavento's books there was a description of his work with a programmer. The programmer literally "burned out" on the job. (He died while working on a neural network.)
This programmer, as I now understand it, was creating a neural network.
Earlier there was such a tool Pesavento map in the Ensign program. Perhaps this tool exists now too.
There was a brief description of this tool in Larry Pesavento's article in the Forex magazine about a year ago.
Pesavento map is that neural network. Here's a bit on the subject here http://ensign.editme.com/pesaventoshow
============
Pesavento's article (in the forex shop magazine and at the link) describes the algorithm of working with the neural network.
In the picture the curve calculated by the neural network is superimposed on the chart.
I recently came across some examples in NS2's helpdesk. Here is one of them. This example, of course, is mostly an NS2 application. But also important is the beginning of this "story", namely what exactly we want to get from the grid and, based on that, how the data for the grid should be prepared.
Just to complete the picture. And also, for reflection.
In one of Larry Pesavento' s books there was a description of his work with a programmer. The programmer literally "burned out" at work. (He died while working on a neural network.)
This programmer, as I now understand it, was creating a neural network.
Happy New Year to everyone, everyone, everyone! Good luck and success to all!
(I don't wish you health. It's just that some people already know where the liver is and some don't know yet... =)))
Here's a link I found. Haven't looked at it myself yet, so I won't comment on the rating, but it's definitely on the topic of this thread.
http://neuroschool.narod.ru/
Alas, the branch is dead. Apparently, the author has created a grail after all. =))
Finally got around to NSH2. It's tight, though. The following question came up: is it possible to import a network from NeuroSolutions 5 into NS2?
Alas, the branch is dead. Apparently, the author has created a grail after all. =))
It's time to pick up a thread, after all it's about NS inputs and I won't start a new thread...
A couple of very simple theoretical questions and thoughts have come up.
*
At). Submitting variants of inputs to the NS, which exaggeratedly look like this:
a) Open[0]-Open[1], Open[0]-Open[2], Open[0]-Open[3], Open[0]-Open[4],..., Open[0]-Open[N]
b) Open[0]-Open[1], Open[1]-Open[2], Open[2]-Open[3], Open[3]-Open[4],..., Open[N-1]-Open[N]
expected to get approximately the same result. The inputs are quite tautological after all. But no, the results are not always even close to each other, both inside and outside the teaching section. I tried different variants of networks and different neural network packs, the picture is almost the same everywhere... Why? Is the learning mechanism imperfect? "Weakness of the network? Is it too much to ask for?
*
Two). I want to hear the opinion of experienced, or consider themselves experienced neural networkers, how far from the price inputs should be in our applied problem of price BP prediction? My experience shows that using all sorts of indicators, "Nth order derivatives" from the price, generally gives nothing at best. Price, maximum smoothing, extremes, everything else from NS opium... Don't you think?