- Search Forex Factory
- 17 Results (16 Replies, 1 Comments)
- Pip_Monk replied Jun 24, 2024
This thread still active? Modern AI's ought to kill this problem. No???
Machine Learning with algoTraderJo
- Pip_Monk replied Jan 22, 2016
I haven't followed this thread for a while but I came across this course at udacity today. Maybe some of you will be interested. url
Machine Learning with algoTraderJo
- Pip_Monk replied Feb 13, 2015
Thanks Felixis. I read their paper and it is a big improvement in NN development. Looks like it simplifies training too. Do you have source code for their method? I code in Delphi (objective pascal) but I can read C and C++. I would really like to ...
Machine Learning with algoTraderJo
- Pip_Monk replied Feb 13, 2015
fromPipMeUp : (I don't understand why the input layer is constrainted. I can use any number of past values to form the input sample.) You're right, but what I mean is the number of inputs is constrained by the number of training features (past ...
Machine Learning with algoTraderJo
- Pip_Monk replied Feb 11, 2015
Hi pipMeUp: Yes, I agree they are awkward. But if they are kept simple they are manageable. In this case, for instance, don't worry about topology, just use the one described by Algo. And use the same activation functions. As a rule, NN's work best ...
Machine Learning with algoTraderJo
- Pip_Monk replied Feb 10, 2015
@beachloup: Thanks for piping up to keep the thread running. I have been following it with interest too. The following quote from AlgoTraderJo may give you what you need (See page 13) : PM.
Machine Learning with algoTraderJo
- Pip_Monk replied Jan 26, 2015
This info will only be of use to members who write their own software. If you want to use the functions in R, such as plotting or NN such as AMORE, you can do a data dump from your software to a file format which can be read into R with the ...
Machine Learning with algoTraderJo
- Pip_Monk replied Dec 16, 2014
Really nice analysis you're doing here algoT. Enjoying the thread a lot. Are you using a stoploss in the recent ensemble models? Are you going to try using an artificial NN as the classifier tools in future? Even though they're a pain to ...
Machine Learning with algoTraderJo
- Pip_Monk replied Aug 30, 2014
I don't understand why the HA is considered important to this system. It leads the 50 ma but is insufficient to trigger a trade. The HA is always "on-colour" by the time a breach occurs, so it doesn't seem to add any new information.
MA of H-A system
- Pip_Monk replied Aug 22, 2014
Thanks for the Transient_Zones indicator FreeFox. Two things I've noticed about the TZ's : One is that the edges of a TZ often coicide with one or more SR levels. The second is that TZ's often form in price ranges where the price doesn't want to ...
The similarity system
- Pip_Monk replied Jun 9, 2014
Good point Chake. I meant a return toward the 50 line. In other words a drop in strength regardless of direction.
A delicate secret
- Pip_Monk replied Jun 9, 2014
Looking at the USDCAD 4H with an RSI(14) it seems that when a candle forms during a drop in the RSI, it has a fairly high probability of being engulfed by subsequent candles in a fairly short time (within one chart page as Eurusdd claims). Of course ...
A delicate secret
- Pip_Monk replied Mar 31, 2012
EMA — The MA is exponential with period of 5 and shift of 3. I have attached a compilation of Big E's posts (word document) which I put together for my own convenience. It contains most of what you need to know. Hope this helps.
Trading Made Simple
- Pip_Monk replied Mar 29, 2012
a picture's worth 1000 words — Thanks for the charts Joffe. Are they 4 hr. ?
Trading Made Simple
- Pip_Monk replied Jan 16, 2012
Great System — EelFranz, thanks for doing this thread. I've been using the methods as outlined in your first few posts with tremendous success. I'm trading the 4H timeframe of the EUR/USD, USD/CAD, GBP/JPY and USD/CHK. I read that you are ill. ...
Trading Made Simple
- Posts by Member Search: 'Pip_Monk'