Machine learning in trading: theory, models, practice and algo-trading - page 69

 
Alexey Burnakov:
Nope. I do not use his handwritten code. I use R and the Expert Advisor for R.

Don't tell me that you're deflated) you can check outside the sample yourself using my data.
The morning at the kolkhoz started. I do not really need it, everything works fine for me. I thought I could help you train your model. Anyway you need a model, you will be able to check it for workability and to announce us all the result, if I generate a model in the optimizer Reshetov????
 
Mihail Marchukajtes:
Well, it's morning at the collective farm. I don't need it, everything works fine for me. I just thought I'd help you train your model. Anyway, you need a model, you can check its workability and announce the result to all of us, if I generate a model in Reshetov's optimizer????
I thought you accepted the challenge. So, go all the way to the end.
 
Alexey Burnakov:
It seems like you took the shuttle. Then bring it to the end.
What do you mean? I'll train a model that will just lie around and no one will use it, I'll waste my machine time, the whole night the computer will rattle in my ear, and the model will just lie around and not needed. Why would I want to do that???? This is bullshit to be honest ..... everything works fine with me guys, I just wanted you to demonstrate the capabilities of Reshetov's optimizer, and if it is not required, then I certainly do not need it ... Everything works fine for me and I am fully aware of what the network can and cannot.... So it's a deal-breaker....
 
Mihail Marchukajtes:
What do you mean? I'll train a model that will just lay around and no one will use it, I'll waste machine time, all night the computer will rattle in my ear, and the model will just lay around and not be used. Why would I want to do that???? This is bullshit to be honest ..... everything works fine with me guys, I just wanted you to demonstrate the capabilities of Reshetov's optimizer, and if it is not required, then I certainly do not need it ... Everything works fine for me and I am fully aware of what the network can and cannot.... So it's a cakewalk....
So show me. We don't have the infrastructure for his software. You go ahead and finish your bragging experiment. And if you do, just say: "the problem is difficult for me, my model is strange, I'm afraid of out-of-sample testing." That's it. We'll get over it. )
 
Alexey Burnakov:
So demonstrate. We don't have the infrastructure for his software. You bring your bragging experiment to an end. If you do, just say: "The task is difficult for me, my model is strange, I'm afraid of out-of-sample testing. That's it. We'll get over it. )
Oh, I see. Well, consider me deflated... And you can safely continue to consider Reshetov's optimizer bullshit, I'm a worthless snob. I'll get over it somehow :-) No one died of profit yet :-)
 
Mihail Marchukajtes:
I see. Well, consider me deflated... And you can safely continue to consider Reshetov's optimizer as bullshit, I'm a useless zany. I'll survive somehow :-) No one died of profit yet :-)
So I would have made the model and even wrote an advisor for it, if not a lot of data, which will be optimized a day or even more..... And during this time, I will need the optimizer to check the signals on the tirnary model, so I don`t need it ..... Although, your file gave me an idea... I'm busy now with construction of my house and it's not really time for trading and definitely not for optimization, but I'll set up my system with about 20 data. According to my estimations 20 entries can saw 400 records, taking into account that I have about 5 signals per day, it turns out 80 days, about 4 months on 5 minutes, where the model should work for about a month or two on 5 minutes, those who understand trading appreciate that this is a fairly long interval for the strategy on 5 minutes, and then put everything in its place. So wait, the result will certainly publish in this thread, but not before next week......
 
Mihail Marchukajtes:
I see. Well, consider me deflated... And you can safely continue to consider Reshetov's optimizer as bullshit, I'm a useless zany. I'll survive somehow :-) No one died of profit yet :-)
Well, we'll get over it. And Yuri all the more so. )
 
Mihail Marchukajtes:
So, I would make a model and even write an advisor for it, if not a lot of data, which will be optimized for a day, or even more..... And during this time, I will need the optimizer to check the signals on the tirnary model, so I don't need it ..... Although, your file gave me an idea... I'm busy now with construction of my house and it's not really time for trading and definitely not for optimization, but I'll set up my system with about 20 data. According to my estimations 20 entries can saw 400 records, taking into account that I have about 5 signals per day, it turns out 80 days, about 4 months on 5 minutes, where the model should work for about a month or two on 5 minutes, those who understand trading appreciate that this is a fairly long interval for the strategy on 5 minutes, and then put everything in its place. So wait, the result will certainly publish in this thread, but not before next week......
Ok. We wait.
 
By the way, I posted the data . Maybe someone will try to get a hold of it? Separately, I'll post a set of out-of-sample to evaluate the trade. There instead of -1;0;1 will be values of differences between prices with an interval of 3 hours. And you can calculate the expectation of the trade on the predicted signals.
 
Mihail Marchukajtes:
Listen! Nobody uses Reshetov classifier here except you, most of us use R programming environment, it's much more flexible approach in all directions than to use some separate product... If you explained properly what and how, then I think each of us would be able to implement both algorithm and trade back test, you know? Just explain properly what to do and how to do it. I told you that a week ago I implemented the same concept, but it did not work, you say it's all bullshit, you need to properly prepare the data, so how to prepare them I ask for the third time? and you will have an implementation and a backtest, and in different variants from each of us....