Machine learning in trading: theory, models, practice and algo-trading - page 276
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
The thing is, I don't want to know about all sorts of "magic" tricks, different frameworks, of which there are 100500, different secret key combinations and what the parameters of one of >10 000 "universal functions" mean. I have a different brain. I myself wrote many hundreds of functions, maybe more than a thousand, and sometimes rewrote some of them many times because I forgot I had already written them, I do not remember the names and signatures of the functions I wrote myself half a year ago and rarely used, how can I remember 10,000 of the left framework? But I'm good at remembering or re-thinking the essence of algorithms, for example on this Fit01 even if I forgot it, I'd have it ready in a minute and it does not depend on OS, the NP, the framework and packages.
R has either 120,000 or 150,000 functions grouped in 10,000 packages.
But, those who use R don't try to find any functions from R which are not in MKL and take them out of laziness or lack of knowledge and use them. This is not the question at all.
The question is posed differently.
For example.
We identify the trading problem: the non-stationarity of financial series.
Then we look in R for tools to solve this PROBLEM.
On this branch a number of problems of TRADING were identified and solutions for them were suggested at the code level.
And the fact that there was a question about the scaling function, it's just working points....
R has either 120,000 or 150,000 functions grouped in 10,000 packages.
But, those who use R do not try to find any functions from R that are not in MKL and take out of laziness, or lack of knowledge and use them. This is not the question at all.
The question is posed differently.
For example.
We identify the trading problem: the non-stationarity of financial series.
Then we look in R for tools to solve this PROBLEM.
On this branch a number of problems of TRADING were identified and solutions for them were suggested at the code level.
And the fact that the question about the scaling function arose, it's just working moments....
R has either 120,000 or 150,000 functions grouped in 10,000 packages.
But, those who use R do not try to find any functions from R that are not in MKL and take out of laziness, or lack of knowledge and use them. This is not the question at all.
The question is posed differently.
For example.
We identify the trading problem: the non-stationarity of financial series.
Then we look in R for tools to solve this PROBLEM.
On this branch a number of problems of TRADING were identified and solutions for them were suggested at the code level.
And the fact that the question about the scaling function arose, it's just working moments....
As for me, then you do not need to look for anything, you need to be able to do yourself. Looking for tools and pulling parameters is a shameful thing to do, not a man's thing.
Words not of a young man but of a husband...
This way of course has the right to life, especially since the R language allows one and the same action to perform in many ways. It all depends on experience, skill level and of course the peculiarities of the "brain". Many people know what it is, like appendicitis, but do not know what for.
PS: I would suggest that the developers to introduce a timer on the forum for self-destruction of records on a timer. As in private correspondence.
That is, here I would put: self-destruction after 2 hours. And would not litter branch huskies.
Good luck
You should be able to do it yourself
It is necessary without conditionally, Zhenya, no one argues, but before you program idea you have to get it from somewhere, I mean a working idea.
There are millions of ideas on the market, but only a few working ideas.
To find a working idea it is necessary to search for it, it is logical?
I don't think you have to search for it.
So you are wrong.
Looking for tools and pulling parameters is a shameful thing to do, not a man's thing.
Of course, it's a hell of a thing here :)
Moving on, you can look for the idea in different ways:
1) to program all by himself ( on a manly, not shameful :)) )
2) Take a ready-made, the very thing that you want but already implemented by someone else.
Since I am not an academic, mathematician or programmer, I chose the second method, since I had no much choice.
I checked everything I wanted, any ideas from cointegrated asset arbitrage to neural networks and hidden Markov models and all sorts of Forrest, as well as technical analysis, levels, candlesticks, spectral analysis, etc.
The picture shows my scripts, written somewhere + - for half a year, each blue file is a script is an idea, in each folder from 5 to 25 such blue files (ideas)
Each script uses from 1 to 6 ready-made libraries (packages)
Each package is thousands or maybe millions of lines of code, often written in a low level language, for speed
So Zhenya summarize, what would you do what I did but in a masculine way :) Ifyou want to do it yourself, you need:
1) be a specialist in many areas of calculus. , statistics, theoretical verification. , digital signal processing, combinatorics, machine learning. and so on and so on.
2) Spend weeks or even months on writing and tuning each of the packages
3) To be able to write in many languages including the old and low-level languages.
4) have 30 years of time to learn and implement all this and the saddest is to realize that it doesn't work :))))
I've spent only half year :) What's better, use your common sense ormanly????
By the way I found what I was looking for, so it was worth the time, I presented my algorithm on page 269
What's interesting, at round levels they mostly put limits, and they put stops based on price levels of the chart.
In crypto it's the same (in terms of limits, stops are not seen there)
In general, wherever there is a deep glass it can be seen.
Of course, I accidentally found a deal in my diary from 5 years ago, the time when I was trading on "oanda". It was a good time, full of hope and some kind of market romanticism. :)
If you have an understanding of what is savings / allocation, understanding who you take the money from, and most importantly discipline and patience, not only to enter successfully, but also to wait out the entire movement, with the last points I have chronic problems, and that's why I joined the algorithms :)
P.S. the whole trading system in one post FREE :))
Nice picture, I like it )
It is necessary without conditionally, Zhenya, no one argues, but before you program idea you have to get it from somewhere, I mean a working idea.
There are millions of ideas on the market, but only a few working ideas.
To find a working idea it is necessary to search for it, it is logical?
So you are wrong.
Of course, it's a hell of a thing here :)
Moving on, you can look for the idea in different ways:
1) to program all by himself ( on a manly, not shameful :)) )
2) Take a ready-made, the very thing that you want but already implemented by someone else.
Since I am not an academic, mathematician or programmer, I chose the second method, since I had no much choice.
I checked everything I wanted, any ideas from cointegrated asset arbitrage to neural networks and hidden Markov models and all sorts of Forrest, as well as technical analysis, levels, candlesticks, spectral analysis, etc.
The picture shows my scripts, written somewhere + - for half a year, each blue file is a script is an idea, in each folder from 5 to 25 such blue files (ideas)
Each script uses from 1 to 6 ready-made libraries (packages)
Each package is thousands or maybe millions of lines of code, often written in a low level language, for speed
So Zhenya summarize, what would you do what I did but in a manly way :) Ifyou want to do it yourself, you need:
1) be a specialist in many areas of calculus. , statistics, theoretical verification. , digital signal processing, combinatorics, machine learning. and so on and so on.
2) Spend weeks or even months on writing and tuning each of the packages
3) To be able to write in many languages including the old and low-level languages.
4) have 30 years of time to learn and implement all this and the saddest is to realize that it doesn't work :))))
I had only half year :) What's better, use your common sense ormanly????
By the way I found what I was looking for, so it was worth the time, my algorithm's work I posted on page 269
Unfortunately all this "high-level" fiddling with 120 000 functions does more harm than help, the main problem is that when you don't know how an algorithm is set up on a low level, and you have to write it yourself at least once without any tips, you won't be able to masterfully use it, modify, improve, create new ones based on it, etc. And everything that is available in the public domain will not give an advantage, especially on the market. So play with 100500 packages based on superficial high-level descriptions, everyone will be happy about it))
Unfortunately, all this "high-level" fiddling with 120,000 functions does more harm than good,
It depends, when a person is looking for something that works, ready packages are better, and when you have already found it, you have to figure it out and rewrite the code yourself
don't you agree?
the main problem is that when you don't know yourself how to make a low-level algorithm, and you have to write it yourself at least once, you won't be able to masterfully use it, modify it, improve it, create new ones based on it and so on
fully agree
all that is in the public domain will not give an advantage, especially on the market.
It's a bit of a blurhere too.
So play with 100500 packs, on the basis of superficial high-level descriptions, everyone will be happy about it)))
ok, whatever :)