Discussion of article "Brute force approach to pattern search (Part IV): Minimal functionality"
one can argue about the value of midprice as (bid+ask)/2, they are both (bid,ask) strictly speaking artificial. And last missing
valuation (bid+1/bid)/N is closer to the golden mean.
It seems to me that in many cases we can consider that the quotation is already a logarithmic chart.
PS/ Here Rena once said about "addition is equivalent to multiplication". perhaps he is right.
one can argue about the value of midprice as (bid+ask)/2, they are both (bid,ask) strictly speaking artificial. And last is absent
valuation (bid+1/bid)/N is closer to the golden mean.
It seems to me that in many cases we can consider that the quote is already a logarithm chart.
PS/ Here Rena once said about "addition is equivalent to multiplication". perhaps he is right.
+1 point more fairly can be in points where the spread has very low values, and then I have not checked, and it is impossible to check. I think it can be different for all brokers, some brokers have +1 and some have +5. There is no benchmark, that's the point, but I have often observed how the stack on Friday falls apart, and it falls apart exactly relative to some average. Part of this physics works for other days of the week too, spreads widen there, and so does the stack. Bruteforce catches mostly these things. It used to catch spread expansions and interpreted them as price movements, after the introduction of this filter everything became much easier and spreads were no longer caught.
Fixed the bugs that I promised to fix:
- Always tickDestroy Spread Noize on the first tab for now ( small bug, I'll fix it later).
- For now, do not set Equation Deep higher than"1" ( also a bug, which I will fix, but in essence you lose nothing, degrees do not give much in terms of results).
If you downloaded the old archive here is the new one with the new version. Now all functions are working, there should be no hangs ( on the second tab could occur )
It's not clear what all this is for? To make a £107 profit on a ten year run in the pound tester? Or maybe go to the square and ask for a handout?
It is interesting you count, not the annual interest rate, but just pounds, and only one currency pair. I'm not even talking about future performance forecasts. You are given a tool for market research, for creating and finding profitable trading systems without human participation, and you are not only too lazy to press the button, you are too lazy to even look into it. I suspect that if I start a new series of articles on neural networks, the essence of your remarks will not change. Now I will ask you what per cent per annum do you expect when trading or using an automatic trading system?
I have downloaded the attachments of the article. I learned to use "Forex Awaiter" to perform brute force search by watching Youtobur’s video. "Awaiter" finally generated the "GBPUSD 5 MATH_WAITING OPTIMIZED.txt" document, but how do I use this document? Or how can this document be turned into a configuration file like *.set?
Also, is this document generated for backtesting through "Reciever" or "Sacred fruit"? Still the same question, how to convert the generated result into the relevant EA setting file?
Thank you very much for sharing, thank you!
I have downloaded the attachments of the article. I learned to use "Forex Awaiter" to perform brute force search by watching Youtobur’s video. "Awaiter" finally generated the "GBPUSD 5 MATH_WAITING OPTIMIZED.txt" document, but how do I use this document? Or how can this document be turned into a configuration file like *.set?
Also, is this document generated for backtesting through "Reciever" or "Sacred fruit"? Still the same question, how to convert the generated result into the relevant EA setting file?
Thank you very much for sharing, thank you!
Thank you very much for your work!
Unfortunately, I do not understand exactly how I convert the data from the awaiter into an MT5 EA
If I understand everything correctly, I have to upload the data from the awaiter (see below) to the Reciever.ex5, right?
Additionally, If you could upload the Reciever.ex5 as .mql5 file, I would be more likely to understand how the Reciever accesses the file from the Awaiter.
Thank you very much for your work! Unfortunately, I don't understand exactly how I convert the data from Anciter to MT5 EA If I understand everything correctly, I should upload the data from the avaiter reciever.ex5, right?
If you can upload the Reciever.ex5 file as a .mql5 file, I would be more likely to understand how to attach to the file from avaiter.Thank you very much! :)
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
You agree to website policy and terms of use
New article Brute force approach to pattern search (Part IV): Minimal functionality has been published:
The article presents an improved brute force version, based on the goals set in the previous article. I will try to cover this topic as broadly as possible using Expert Advisors with settings obtained using this method. A new program version is attached to this article.
The problem with many automated trading systems is that they get overtrained and overfit to history. It is possible to create a system that will show impressive results, up to 1000 percent per month. But such systems do not work in reality.
The more input parameters a trading system has and the greater the variability of the EA logic, the stronger such an EA sticks to history. The thing is that we have a very simple process of how a quote is converted to another data format. There are always forward and backward conversion functions that can provide both forward and backward data conversion process. It can be compared to encryption and decryption. For example, WinRar archive is an example of encryption. In the context of our task, the encryption algorithm is a combination of the optimization process and the presence of trading logic. A sufficient number of backtests in the optimizer and a certain flexible logic can work a miracle. In this case, the trading logic serves as a decoder that decodes future prices based on the readings of the past.
These settings are attached below, so you can test them if you wish. You can also try to find your own settings and backtest them on demo accounts. Starting with this version of the program, anyone can try this method.
Author: Evgeniy Ilin