Backtesting/Optimization - page 93
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
They will not correct that any time soon
It was working, then it was not, then again it was working, and after the last time it does not works for the last 4-5 months. they do not have an intention to correct it (I have sent a message to support desk and it was ignored)
I am just going to tell this at this occasion :
According to metaquotes, back test optimization is using genetic algorithm. Now, just two usual sentences about genetic algorithm :
1. The evolution usually starts from a population of randomly generated individuals
2. Commonly, the algorithm terminates when either a maximum number of generations has been produced, or a satisfactory fitness level has been reached for the population.
The bolded text means one thing, and one thing only as far as back test is concerned : due to the randomness nature of genetic algorithm, it is impossible to have same results in two consecutive EA optimization tests even with same parameters (there is no fitness criteria anywhere - the "limitations" are anything but it). Now do two optimization tests (or 100 if you want) with same parameters on same data using the "genetic algorithm" and see what are you going to get
_________________________
Number 2 : how in the name of good are consecutive tests of the same EA on the same data with same parameters always the same? They do not have ticks data. Which excludes same results. Which means that they were not even able to make some sort of pseudo random ticks generation but are always doing same "pseudo ticks" generation which makes back testing literally worthless (those are "metaquotes ticks" which never change, not ticks, and can not simulate worst case nor best case scenarios)
_________________________
Combining the two : using big words to hide a fact that the worth of something is close to nothing is a well known way we are facing each and every day on internet when we see some wonder indicators or EAs being sold. So much about the whole thing (I could talk for hours about metatrader back testing but there is no point to do so - it will not change the fact that it is a simple snake oil selling and not back testing)
I am just going to tell this at this occasion :
According to metaquotes, back test optimization is using genetic algorithm. Now, just two usual sentences about genetic algorithm :
1. The evolution usually starts from a population of randomly generated individuals
2. Commonly, the algorithm terminates when either a maximum number of generations has been produced, or a satisfactory fitness level has been reached for the population.
The bolded text means one thing, and one thing only as far as back test is concerned : due to the randomness nature of genetic algorithm, it is impossible to have same results in two consecutive EA optimization tests even with same parameters (there is no fitness criteria anywhere - the "limitations" are anything but it). Now do two optimization tests (or 100 if you want) with same parameters on same data using the "genetic algorithm" and see what are you going to get
_________________________
Number 2 : how in the name of good are consecutive tests of the same EA on the same data with same parameters always the same? They do not have ticks data. Which excludes same results. Which means that they were not even able to make some sort of pseudo random ticks generation but are always doing same "pseudo ticks" generation which makes back testing literally worthless (those are "metaquotes ticks" which never change, not ticks, and can not simulate worst case nor best case scenarios)
_________________________
Combining the two : using big words to hide a fact that the worth of something is close to nothing is a well known way we are facing each and every day on internet when we see some wonder indicators or EAs being sold. So much about the whole thing (I could talk for hours about metatrader back testing but there is no point to do so - it will not change the fact that it is a simple snake oil selling and not back testing)Dear SIR MLADEN,
In this case, the tester is best in used just to see if our newly created Indicator or EA is working ( not triggering any errors)... that's all...
The result produce by it would not be reliable... have to forward test and tweak it along the way.... as recommended SIR
yours truly
AZRUL...
Hi coders,
here it is what im looking for:
given a symbol & timeframe chart where i choose an indi, how to set its best parameters to avoid the false signals ?
Is there already a tool (ea/script) that only return the best list of possible parameters (instead of using the default ones) ?
This exclude the 'metatester' as it works in a very different way. Here the found settings (computed on the past data) need to be confirmed by the next data (forward checking) to validate the found settings.
Thanks to share if you know about such a tool.
Hi coders,
here it is what im looking for:
given a symbol & timeframe chart where i choose an indi, how to set its best parameters to avoid the false signals ?
Is there already a tool (ea/script) that only return the best list of possible parameters (instead of using the default ones) ?
This exclude the 'metatester' as it works in a very different way. Here the found settings (computed on the past data) need to be confirmed by the next data (forward checking) to validate the found settings.
Thanks to share if you know about such a tool.dino35
There is no such thing
dino35 There is no such thing
Thanks mladen,
maybe can you point me to some 'auto-optimization' indi's criteria model (sort of fuzzy logic inside)
Hi Mladen, and Mrtools,
Can you help optimize this great indi ?
Every-time I set extern bool ShowBreakup =1 my CPU ( i7 3770k) goes crazy ( +65 Celsius, 70%+ utilization ). It does not freeze , but it's really,really slow,can't trade with it ...
Thanks !
Hi Mladen, and Mrtools,
Can you help optimize this great indi ?
Every-time I set extern bool ShowBreakup =1 my CPU ( i7 3770k) goes crazy ( +65 Celsius, 70%+ utilization ). It does not freeze , but it's really,really slow,can't trade with it ...
Thanks !razo
Try it out now : dt_zz_supres_fast.mq4
razo Try it out now : dt_zz_supres_fast.mq4
Mladen,you're trully good,thank you Sir ! Beers up !
razo Try it out now : dt_zz_supres_fast.mq4
Zdravo Mladen, would you be kind enough to add an alert to it ? Sound + PopUp indicating which pair is acting up? If you can please get it to buzz only on the first dot of the opposite color ( when new leg is formed ) would be perfect !
Many thanks !