_Market description - page 25

New comment
 
Figar0 >> :

Such an answer is inevitable. Except that points 4 and 5 say that one cannot work without multi-currency analysis, i.e. as a conclusion, experts without such analysis are not workers either. Interesting...

Let's see if someone else tries to predict something, maybe what we have is enough for someone else.

down

up

up

 
dasmen писал(а) >>
Prival 14.03.2009 10:39 PM Maschka measures the ratio of price changes over time - now it's all there, and here the parameter itself in the context of code in search of a winning MA, although you can just call it effective, and the accuracy in this case is calculated by the ratio of the muv period itself to the number of bars and the number of trends that is directed muv movements with a given period, which muv forms at a given range of measurements, well, put it this way: WiningMA = NumberOfBars/(NumberOfTrends*PeriodMA) And the coefficient of reliability of the method is considered as follows: kWiningMA = NumberOfBars/PeriodMA When studying the same statistics I did not meet the use of this method and the choice of names is arbitrary, well let it be called like this.

the curve you see on the screen does not have this property of winning or losing. the ct you made up has this property, but the curve does not have it....

 
Prival 14.03.2009 15:07 Well, in general the definition of gain or loss is the area of definition of the same statistics, here the method uses selection and for the operations of selection it is a well-established contextual definition, what is the contradiction? The MA itself in this context is a measurement tool that participates in the method, and the method itself has this definition and it is standard, the period or the number of bars are properties, and in general these are all truths that are commonplace. And in the same case when we measure the same voltage with the same voltmeter - it all looks exactly the same, because the error of the same measurements when measuring - this is also not an end in itself, but part of the method for identifying something and in one case may vary and the range of possible deviations in general is revealed simply by experience. This is not a TS, but just a selection method for identifying individual statistical properties of the market, that is, to identify the best MA periods to use such a selection method just think more effective way For example there is the same neural network and there is the same problem, what data to use to train it, so this scheme allows us to first identify the actual range from which we will feed the NS, in the other case it may be a system at the intersection of MA - this is all derivative. I still think that there are not so many of such recipes in the entire Forum - more and more people are looking for them.
 
dasmen писал(а) >>
Prival 14.03.2009 15:07 Well, in general the definition of winning or losing is the area of definition of the same statistics, here the method uses selection and for in the operations of selection is a well-established contextual definition, what is the contradiction? The MA itself in this context is a measurement tool that participates in the method, and the method itself has this definition and it is standard, the period or the number of bars are properties, and in general it is all truths that are commonplace. And in the same case, when we measure the same voltage with the same voltmeter - it all looks exactly the same, because the error of the same measurements during measurements is also not an end in itself, but are part of the method to detect something and in one case can vary and the range of possible deviations in general to a greater extent is revealed simply by experience. This is not a TS, but just a selection method for detecting some statistical properties of the market, i.e. to detect the best periods of MA to be used by such a selection simply consider it more effective than trying MA parameters already integrated into TS, after that it is easy to use in construction of TS on the best of obtained parameters. For example, there are the same neural networks and there is the same problem - what data to use for training, such a scheme allows us to detect the actual range from which to feed the NS first, in the other case it can be a system based on intersection of MA - all this is a derivative. I still think that there are not so many of such recipes in the entire Forum - more and more people are looking for them.

I understand what you are saying. I wish you could have understood me.

Z.I. By the way, the best MA periods can be found in another way, without going over it.

 
Prival 16.03.2009 00:21 Not getting it is unlikely, I'll respond with an excerpt from the same Statistics on time series analysis: "There are two main purposes of time series analysis: (1) determining the nature of the series and (2) predicting (predicting future values of the time series from present and past values). Both of these purposes require that the model of the series be identified and, more or less, formally described. Once the model is defined, you can use it to interpret the data in question (e.g. use it in your theory to understand seasonal changes in commodity prices, if you are dealing with economics). Without regard to the depth of understanding and fairness of the theory, you can then extrapolate the series based on the model found, i.e. predict its future values." - Based on the above, there is identification of seasonal and trend-cyclical, again according to statistics and since seasonality in the Forex market is not identified and the value of the trend cycle is not fixed, the average cycle itself is determined by the above technique - that is important and the MA period is not an end in itself. You just want to find an answer that confirms the opinion you have and in the way you wish, but that is not an obligation for me at all.
 
Prival писал(а) >>

1. upwards

2. upwards

3. from the 5th bar pullback

although there is too little info, just an attempt to guess no more than that

It is good that it is missing). Maybe if we understand exactly what is missing in this picture, we will understand what to take into account. If not difficult in a nutshell, what is missing that was not a guessing game, and something at least somewhat meaningful? I think what we are missing is the TF and the tool.

 
Oh, by the way, who hasn't played the "interesting" game yet, the picture on page 23 is waiting for your answers. We need to get some more statistics. Yes, and you can write right away some of the most important points that are missing for a more meaningful prediction, I wonder how individual it is.
 
Figar0 писал(а) >>

It's good that it's missing). Maybe by understanding exactly what's missing from this picture, we'll understand what exactly needs to be taken into account. If it's not too difficult, in a nutshell, what is missing so that it would not be a guessing game, but something at least a little bit meaningful? I think the main thing that is missing is the TF and the tool.

For me personally, first of all the grid. 'Grid'.

Very good idea by the way, maybe that's the way to go.

 
Trading real money on forex is like having sex with teenagers:
- Everyone thinks about it;
- everybody talks about it;
- everybody thinks their neighbour is doing it;
- almost no one does it;
- whoever is doing it is doing it badly;
- everybody thinks they'll do it better next time;
- no one takes safety precautions;
- anyone is ashamed to admit they don't know something;
- if someone succeeds at something, there is always a lot of fuss about it.
 
- Why do you work on FOREX?
- Well, you can work on FOREX in three cases. One: if you have ability and money. The second: if there is no money, but there is ability. Third: if there is no ability, but there is money.
1...1819202122232425262728293031
New comment