- Search Forex Factory
- 11 Results
- emeafx replied Dec 28, 2015
Hi Guys, i would do some detailed analysis of this indi, but first need a version written in "R". If someone have it would save me time. Not need to be rewritten.
- emeafx replied Dec 21, 2015
Lebanese nice work! Could you tell us something about the indicator itself? I saw the source but maybe it is better if you shed some light... :-) THX
- emeafx replied Jan 21, 2015
OK, got it. :-) What about combining faster reacting a slower reacting systems (maybe not 2 but more) to an ensemble. My biggest concern to avoid long deep DD periods even if it makes lower (but steadier ) returns. Ideally the ensemble should ...
- emeafx replied Jan 21, 2015
Regarding ensemble: there is a strong correlation between the two equity curve, so it (the ensemble) can not reduce the drawdown significantly. My idea: if we could find two NN algorithm with not so strict correlation than the ensemble could ...
- emeafx replied Jan 19, 2015
Did you try some weighting between the two algorithms based on the past performance? I mean if NN had good perfomance it has "more votes" regardind the decision on whether going into a trade or not.
- emeafx replied Jan 8, 2015
Jo, where do you get your data (maybe Dukas)? I would make some stat. analysis on it. Or would you mind to share it?
- emeafx replied Jan 7, 2015
These are very nice curves, but for me a little bit disturbing the long (6months?) heavy drawdown period. Did anyone try to investigate (some papers maybe) what property of the time series changing causing this drawdowns? I tried several change ...
- emeafx replied Dec 23, 2014
Thanks! I will look it!
- emeafx replied Dec 23, 2014
sorry for being unclear. How can you train a linreg model on 2 inputs? For my understanding the linreg takes the inputs and gives you back a slope and intercept what does not depends on your target. If it is an NN I understand the process.
- emeafx replied Dec 23, 2014
Hi Jo, first of all Merry Xmas for all of you here in the forum! some remarks: did you try Theil-Sen estimator instead of linear regression (there is good R package "Zyp") I try to reproduce your number in full R, could you define the algorithm ...
- emeafx replied Apr 10, 2014
Hi, could anyone shed some light why this pair cannot stupidly going higher? thx
- Posts by Member Search: 'emeafx'