Machine learning in trading: theory, models, practice and algo-trading - page 226

 
Konow's tag:
If we're talking about 1000 or even 3000 thousand lines of code, it's not a problem. An experienced developer can do it. Just need to understand the essence of this MO and write it. I would do it, but I'm busy with graphics at the moment. Maybe there are some enthusiasts?

We're all busy...

How about dropping everything and demonstrating by deed how it would look beautiful with the randomforest algorithm on the µl. Specifically. At least one algorithm, the most common in machine learning. At the same time, you can help the metaquotes too...

 
Tag Konow:
If we're talking about 1000 or even 3000 thousand lines of code, it's not a problem. An experienced developer can do it. You need to figure out the essence of this MO and write it. I would do it, but I'm busy with graphics at the moment. Maybe there are some enthusiasts?
I like your enthusiasm, but MO is not riveting graphics, it is difficult, you need to be clever, for example MLP with C++ backend is about 100-150 lines of code, but all mankind invented backend for 20 years, which in general project no more than 30-50 lines of code, how some zero can create 10000 lines of code in Dreamviewer without getting into code with the help of visual tools in a couple of hours, so code is not like code. there were 200 pages long threads about neural networks which ended up with nothing
 
SanSanych Fomenko:

We're all busy...

How about dropping everything and demonstrating by deed how it would look beautiful with the randomforest algorithm on the µl. Specifically. At least one algorithm, the most common in machine learning. At the same time you can help the metaquotes...

Let's put it this way:

You are a follower of R and MO, and therefore you understand both.

You explain to me the essence and principles of MO, and I'll tell you if I can write it and how long it will take.

Based on this, it will be possible to find free time and start implementation. Perhaps some helpers will be needed.

 
Tag Konow:

Let's put it this way:

You are a follower of R and MO, and therefore you understand both.

You explain to me the essence and principles of the MO and I will tell you if I can write it and how long it will take.

Based on this, you can find free time and begin implementation. Perhaps some helpers will be needed.

MO (probably "machine learning") is thousands of algorithms. Which one do you want?
 
toxic:
I like your enthusiasm, but the MO is not riveting graphics, it is difficult, you need to be smart, for example MLP with C++ backprops about 100-150 lines of code, but all mankind invented backprops for 20 years, which in a common project no more than 30-50 lines of code, how some zerobie in Dreamviewer without getting into code using visual tools can produce 10000 lines of code in a couple of hours, so code is not the same thing, there were 200 pages long threads about neural networks which ended up with nothing

Obviously, the problem is in the formulation of the essence. You really have to be smart for that).

Formulate the essence accurately. That's enough to assess the complexity of the problem.

By the way, quality graphics is the hardest thing, in case you didn't know.

 
SanSanych Fomenko:
MO (probably "machine learning") is thousands of algorithms. Which one do you want?
Please give me the gist of it.
 
Tag Konow:

Obviously, the problem is formulating the point. You really have to be smart to do that.)

Formulate the gist accurately. That's enough to assess the difficulty of the task.

By the way, quality graphics are the hardest thing to do, in case you didn't know.

That's like asking "formulate the essence of higher mathematics", go to a tech school in 3-5 years you will be formulated in general terms.machine learning 2014 here look at the basics
 
toxic:
This is like asking "formulate the essence of higher mathematics", go to a technical university in 3-5 years you will be formulated in general terms.machine learning 2014 here look at the basics

Well, the essence can be formulated in a set of sentences. That's all I ask.

Now the task is not to study the topic, but to preliminarily estimate its scope. That's why I say formulate (if you understand).

 
Tag Konow:
Please give me the gist of it.
The gist, no problem. You take it and do it.
 
SanSanych Fomenko:
The essence, no problem. You take it and do it.

So there are 200 pages of discussion about a topic that no one can articulate? Why do you disappoint me so much?)

I'll wait for the wording and then I'll estimate the labor costs. Then it will become clear whether you can do without R on your own, or not.