Evaluating the effectiveness of filters in the construction of an ATC - page 3
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
Since there are no comments after the last post, there are two assumptions - the topic is not interesting or I don't understand what I'm writing about. So I decided to post a live example in a file, which shows how the data is compared and the best options are chosen.
I will be glad if the topic will be developed.
Aleks, imho, more likely the second. To understand something and maybe argue with something, you need complete raw data. Your Table is the final grouping of data. But the topic is very interesting...
Last time I attached the data (almost the raw data - convolution from the raw data) on the basis of which a decision is made, there are formulas and several selection options.
Have you looked at this file? If you want the raw data of the results of each pass, I can also post it.
A filter, by itself, can be neither effective nor ineffective - it just has some properties. A filter is designed (matched) to a specific strategy. A filter is effective if it is fit for purpose - the buttons are well sewn, no complaints (c). That said, the strategy itself can be three times ineffective.
In general, I don't understand what the topic implies about the effectiveness of filters.
A filter, by itself, can be neither effective nor ineffective - it just has some properties. A filter is designed (matched) to a specific strategy. A filter is effective if it is fit for purpose - the buttons are well sewn, no complaints (c). That said, the strategy itself can be three times ineffective.
All in all, I don't understand what the topic implies about the effectiveness of filters.
The filter is used to solve a certain problem - for example, not to buy when it is expensive, how it copes with this hadache and is the effectiveness, respectively, the theme is designed to understand what criteria to use to evaluate this effectiveness - the change in profitable trades, the change in the recovery factor and so on. I use a number of indicators for evaluation, trying to compare a number of methods in order to determine the more efficient one.
What else is left unclear to you?
The filter is used to solve a specific problem - for example, not to buy when expensive, how it copes with this hadache and is the effectiveness, respectively the topic is designed to understand what criteria to use to evaluate this effectiveness - the change in profitable trades, the change in the recovery factor, and so on. I use a number of indicators for evaluation, trying to compare a number of methods in order to determine the most effective one.
What else remains unclear to you?
You are confusing a filter with a strategy. The task of filtering is to isolate any components from the input sequence, but in no way to transform, much less interpret the results.
Everyone has to do their job. No filter, by definition, is able to tell when it is expensive and when it is cheap - it is none of the filter's business.
You are confusing a filter with a strategy. The task of filtering is to isolate any components from the input sequence, but not to transform, much less interpret the results.
Everyone has a job to do. No filter, by definition, is able to tell when it's expensive and when it's cheap - it's none of the filter's business.
Don't impose your philosophy on others. I have told you the facts, you are trying to understand what is said through the prism of your theory.
Read the first post - I made a clear distinction between strategy and tactics there. Filters refer to tactics, as I understand it.
Judging by what I underlined in your reply, I conclude that you do not understand what I write - read it again, or admit that the author of the thread is incapable of conveying his thoughts to you.
Don't impose your philosophy on others. I have told you the facts, but you are trying to understand what I have said through the prism of your theory.
Read the first post - I made a clear distinction between strategy and tactics there. Filters belong to the tactics, as I see it.
Judging by the emphasis I put on your reply, I conclude that you do not understand what I write - read it again, or admit that the author of the thread is incapable of communicating his thoughts to you.
It's not my theory, it's yours. I use a generally accepted definition of filtering. Understand what it is called and what it is intended for, and it will be easier for you. For a start, at least with the definition of filtration, in the conventional sense.
I come to the conclusion that the author is unable to convey his thoughts to anyone at all. As it is there - I confuse kings with aces, and I confuse duplets with openings.
It's not my theory, it's yours. I am using a common definition of filtration. Figure out what is called and what it is intended for, and it will be easier for you. For a start, at least with the definition of filtration, in the generally accepted sense.
I come to the conclusion that the author is unable to convey his thoughts to anyone at all. As it is there - I confuse kings with aces, and I confuse duplets with openings.
Of course the theory is mine and I'm ready to discuss it here and argue about it substantively. Besides, I'm not averse to other theories, but I don't welcome unsubstantiated reasoning.
You correctly noted that the filtering is used for the selection, separation by attribute - the signal of the strategy is filtered, because besides the basic (an asset goes up or down - depending on the strategy) indicators for market entry, there are additional indicators, designed to examine the signal situation for market entry under a magnifying glass and check the market situation for absence of unfavorable attributes, which statistically lead to the opposite result. That's why filtering is used not for market data itself but for strategy signal filtering which is designed to give an answer not what to do but when to do it.
In addition, applying filters only after receiving a signal from a strategy to enter the market helps to optimise the computational process.
Any filtering reduces the number of trades, hence the effectiveness of the filter must be considered by this parameter. Provided that the filter increases profit and reduces drawdown. Otherwise, the filter is wasted.
So, an effective filter improves trading parameters with a minimal decrease in the number of trades.
Any filtering reduces the number of trades, hence the effectiveness of the filter must be considered by this parameter. Provided that the filter increases profit and reduces drawdown. Otherwise, the filter is wasted.
So, an efficient filter improves trading parameters with the minimum number of trades.
That sounds logical! However, the question is what to measure with? :)
Decreasing the number of trades will inevitably lead to a change in the estimated parameters of a strategy (set), that is the question of how much it is acceptable to change them.
For example, optimization without filtering an average profit of 100k and 35% of losses (loss of deposit) in case of an average number of deals 500, and after using the filter an average profit of 10k and 5% of losses in case of an average number of deals 250 - profit decreased by 10 times and percentage of bad outcomes only by 7 times, whereas the number of deals decreased by 2 times - is it good or bad? Which is the higher priority?