Machine learning in trading: theory, models, practice and algo-trading - page 766
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
I read about econometric models, something new to me:
Over the last few decades, models such as regime switching (RS) have become popular in econometrics. For example, the threshold autoregressive (TAR) model [15] and the smooth transition autoregressive (STAR) model allowing for smooth transition [3] draw people's attention. Although interest in RS models has grown, most papers on RS have focused on model development, with only a few applications of RS models concerned with artificial intelligence being found in the financial trading field.
Have you tried any of these?
I have tried a lot of them.
Package called tsDyn, it has some threshold functions, I used SETAR instead of Bolinger. I recommend it, it's very effective when using increments. I don't know why, it was about 3 years ago.
Tried and lots of them.
The package is called tsDyn, it has several threshold functions, I used SETAR instead of bolinger. I recommend it, it's very effective when using increments. I've never used it for real account, I don't remember why, it was 3 years ago.
I can send you an article, it uses garch on the input and a neural network for continuous optimization of the utility function
I can send you an article, it uses garch on the input and a neural network for continuous optimization of the utility function
If you don't mind
There's a whole area of research on markets, as it turns out
there are a lot of articles on the subject.
I looked it up, thank you! But this topic falls out of what I'm doing, so I won't get into it until the current stuff is done.
As you might say, without any embellishments. In the marked place there was a change in the futures, and the TS was not overtrained and traded hands. The second, all the more it will be a confirmation of the working theory. Will I be able to move the second time with a five-gauge in four times????
Now I'm getting serious about building the model. It turns out not so much because of the gluing of the futures, but I bet that with ease I will repeat the result..... I can feel it... Because today I did half of the data processing and invented another handy gizmo which helps in the work. To balance the output for the same number of zeros and ones (VERY IMPORTANT that the number of zeros and ones in the output was equal), I had to do permutations of the past data. Now I found a way not to do this, I just build a sliding window of sum of ones and choose the output which has the number of ones exactly half the size of the selected window. This does not lose serial link between the data, because no permutations are done, and selected exactly the window at the desired time interval, where the output is already balanced. So, as best I could... explained. If you don't understand, it's not my fault :-)
No extra words, no extra phrases. Friends welcome you!!!!!
What a fucking change in the futures... How convenient, and it came on such important news as the change in the Fed rate..... A confluence of circumstances....
Not to mention the fact that I sat at the terminal and started moving my stops.... Moron :-(
As you might say, without any embellishments. In the marked place there was a change in the futures, and the TS was not overtrained and traded hands. The second, all the more it will be a confirmation of the working theory. Will I be able to move the second time with a five-hundredweight in four times????
Now I'm getting serious about building the model. It turns out not so much because of the gluing of the futures, but I bet that with ease I will repeat the result..... I can feel it... Because today I did half of the data processing and invented another handy gizmo which helps in the work. To balance the output for the same number of zeros and ones (VERY IMPORTANT that the number of zeros and ones in the output was equal), I had to do permutations of the past data. Now I found a way not to do this, I just build a sliding window of sum of ones and choose the output which has the number of ones exactly half the size of the selected window. This does not lose serial link between the data, because no permutations are done, and selected exactly the window at the desired time interval, where the output is already balanced. So, as best I could... explained. If you don't understand, it's not my fault :-)
No extra words, no extra phrases. Friends welcome you!!!!!
If I were you, I would not change anything in the TS, which has shown, by the way, not bad results
Do not interfere with her work, because all thoughts are already in her and the system does not change your mind
let it run purely automatic, no need to spoil it with your handsIf I were you, I would not change anything in the TS, which by the way showed a good result
Do not interfere with her work, all thoughts are in her and the system cannot be changed
Let it run purely automatic, no need to use your handsSo I entered with my hands exactly according to the readings of the TS. If I hadn't moved my stop, I would have gotten two losses. I would have gone down to about 1650 and would have moved on, but no... In the previous screenshot, I showed how it was just a couple of noughts. These are these two elks. If they hadn't been taken out, I would have recovered at once, there was just a rollback to them..... All right. It is more interesting, how we will move on.... :-)