Machine learning in trading: theory, models, practice and algo-trading - page 2889

 
mytarmailS #:

How do you make yourself hold your position longer? Who's struggling with that?

I know how to wait.

I know how to pony up.

I know how to walk in

I don't know how to hold--


The main problem, after entering, there is often a reverse signal...

If you ignore it (keep holding it) it triggers almost always.

If you don't ignore it, the movement develops (you should have held it).


Off-topic, of course, but there's no one to ask.

And from these trades we could have squeezed out obviously more

The most obvious filter for any signal is the same signal on a neighbouring, older timeframe.

 
mytarmailS #:

What's important when analysing the market...What I've learned.....

1) Absolute price is important

2) Levels are there and they are important, but they are not what is written about in books.

3) Price is not SB.

4) The market is made up of participants, like a substance is made up of atoms, and it should be investigated by such methods.

5) The market is quite deterministic.

A few days ago I came across a smooth package that works with prices and therefore with trends. A bunch of smoothing methods with predictions, in all methods lies the idea of state space.

As a joke, it's New Year's Eve after all, an appeal to one of the functions:

Usage
adam(data, model = "ZXZ", lags = c(frequency(data)), orders = list(ar =
c(0), i = c(0), ma = c(0), select = FALSE), constant = FALSE,
formula = NULL, regressors = c("use", "select", "adapt"),
occurrence = c("none", "auto", "fixed", "general", "odds-ratio",
"inverse-odds-ratio", "direct"), distribution = c("default", "dnorm",
"dlaplace", "ds", "dgnorm", "dlnorm", "dinvgauss", "dgamma"),
loss = c("likelihood", "MSE", "MAE", "HAM", "LASSO", "RIDGE", "MSEh",
"TMSE", "GTMSE", "MSCE"), outliers = c("ignore", "use", "select"),
level = 0.99, h = 0, holdout = FALSE, persistence = NULL,
phi = NULL, initial = c("optimal", "backcasting"), arma = NULL,
ic = c("AICc", "AIC", "BIC", "BICc"), bounds = c("usual", "admissible",
"none"), silent = TRUE, ...)

What follows is a 4-page description of the meaning and applications!

These are trend models, i.e. working with price.

smooth: Forecasting Using State Space Models
smooth: Forecasting Using State Space Models
  • cran.r-project.org
Functions implementing Single Source of Error state space models for purposes of time series analysis and forecasting. The package includes ADAM (Svetunkov, 2021, < https://openforecast.org/adam/ >), Exponential Smoothing (Hyndman et al., 2008, < doi:10.1007/978-3-540-71918-2 >), SARIMA (Svetunkov & Boylan, 2019 < doi:10.1080/00207543.2019.1600764 >), Complex Exponential Smoothing (Svetunkov & Kourentzes, 2018, < doi:10.13140/RG.2.2.24986.29123 >), Simple Moving Average (Svetunkov & Petropoulos, 2018 < doi:10.1080/00207543.2017.1380326 >) and several simulation functions. It also allows dealing with intermittent demand based on the iETS framework (Svetunkov & Boylan, 2019, < doi:10.13140/RG.2.2.35897.06242 >).
 
Valeriy Yastremskiy #:

With logarithmic data of course the weights will be different, smaller values will be taken into account less, larger ones, more than in the series of absolute or relative increments or data. These are for different decision problems after all.

That is why wooden models are good: no transformations interfere, and they are not necessary. Both increments and their logarithms and absolute prices will get splits in the same place.
 
СанСаныч Фоменко #:

A few days ago I came across a smooth package that works with prices and therefore with trends. There are a lot of smoothing methods with predictions, all methods are based on the idea of state space.

As a joke, after all, it's New Year's Eve, and one of the functions is called:

This is followed by a 4-page description of the meaning and applications!

These are trend models, i.e. working with price.

I recommend.

To be sure of the quality of the package, check it today in an hour on the real market.

This is the only way to be sure without involving outsiders.

Convincing results shown by you will be a plus to the said post.

 
Uladzimir Izerski #:

I don't think the jokes are appropriate. As CBs are sometimes bought at the highs and sold at the lows.

For speculators the meaning of this action of the Central Bank will not be clear for a very long time) Learn economics.

They can buy anywhere, that's not what we were talking about....

The conversation is about the fact that structures with a global understanding of the situation, with a staff of analysts, with unlimited financial reserves - make a decision on the mashka, on the mashka CARL!!! :)))

that's one!


and two!

Now we need to come down to earth a little bit and understand that we trade inside the day, in the 5th decimal place. Inside a day candle and even inside a week candle there is purely speculative population, no bank reacts to this noise, fluctuations, there is nothing to react to....

Once again, our TF is 5 digits, banks' TF is 5 digits.

count the number of decimal places, that's the TF of the Central Bank and banks.


And to think that the euro rose by 50 pips intraday is because of the central bank? Because of the mashka? Well, it's more than ridiculous.


No offence to anyone, but just think about it and you'll understand...

 
Uladzimir Izerski #:

Recommended.

To be sure of the quality of the package, check today in an hour on the real market.

This is the only way to be sure without involving outsiders.

Convincing results shown by you will be a plus to the said post.

The author has posted a lot of materials on his package, google gives a large list, the link to the package itself also has a rather large list.

If we talk about the reasoning above, the author claims that trends in financial markets start to appear from the days, so trend models should also be used from the days. And he tries to take into account the nuances of trend patterns, which have appeared since the days, and which were also discussed above.

By the way, the problem of outliers was once discussed here. This is how the package works with emissions. I think so.

S

will be a plus to the above post

The problem is that I have never been interested in the "pluses" of my person since childhood. And I advise you the same - to sneeze at the opinion of other people. The judge must sit within the person.

 
mytarmailS #:

They can buy anywhere, that's not what we're talking about....

And the conversation is about the fact that structures with a global understanding of the situation with a staff of analysts, with unlimited financial reserves - make a decision on the mashka, on the mashka KARL!!! :)))

that's one!


and two!

Now we need to come down to earth a little and realise that we are trading inside the day, in the 5th decimal place. Inside a day candle and even inside a week candle there is purely speculative population, no bank reacts to this noise, fluctuations, there is nothing to react to....

Once again, our TF is a 5-sign, the banks' TFs are

Count the number of decimal places, that's the TF of the Central Bank and banks.


And to think that the euro rose by 50 pips intraday is because of Central Bank? Because of Mashka? Well, that's more than ridiculous.


No offence to anyone, but just think about it and you'll understand.

You don't open the news for me, because I pointed out to you the realities that already exist. Same thing.)

 
Aleksey Vyazmikin #:

Above you write that the state through the Central Bank influences the trades to a greater extent, and here you write that matstat and MO rules - are you really sure that the Central Bank does not use technical analysis?

A couple of years ago I posted a "methodology" from the Central Bank, where it was stated what "indicators" are recommended to use for decision making, and back in 2014 the head of the Central Bank of Russia spoke about Mashka 360 as a measure for making trading decisions.

To recommend does not mean to use)

The moving average, although used by visualisers, is not their invention. It has been used since the matstat appeared. And visualists invented all the things that can be seen but poorly formalised - waves, for example, or support lines and so on and so forth.

The moving average is easy to calculate and formalise, and the work of visualists on its basis is, for example, the "rainbow fan" of averages in some forum thread).

 
Uladzimir Izerski #:

check it out in an hour today in the real market.

I was torn up on the news)) three orders stopped in a second)) I forgot to take them off))))

 
Aleksey Nikolayev #:

To recommend doesn't mean to use)

The moving average, although used by visualisers, is not their invention. It has been used since the matstat appeared. And visualists invented everything that can be seen, but poorly formalised - waves, for example, or support lines and so on and so forth.

The moving average is easy to calculate and formalise, and the work of visualists on its basis is, for example, the "rainbow fan" of averages in some forum thread).

Well, I probably wouldn't move matstat away from technical analysis, technical analysis is rather a visualisation of mat analysis. Whether there is validity of technical analysis is not important, but what is important is that it serves for decision making in different banks, mutual funds and funds, as people are taught it and these people with profile education go to work there, and they should justify their actions clearly for external users.

There are justified levels - for example, trading limits, when reaching which the trading session stops, then option strikes. In the Russian Federation, there is also a budget rule for buying/selling currency, which is implemented relative to the oil price. There are probably levels at which the Central Bank considers the exchange rate unreasonable and is ready to interfere in trading, which they announce on their website. There are natural levels, when a large participant shorted a paper under REPO and now does not want to get into a loss and protects the position.... there's a lot of things.

By the way, who calculates volatility in their indicators as the Central Bank? :)

And yes, I read that the Central Bank uses IMF methodologies.

Reason: