Discussing the article: "MQL5 Wizard Techniques you should know (Part 29): Continuation on Learning Rates with MLPs"
Dear Stephen,
Thank you for sharing your knowledge and work regarding this trading system!
I followed your article with great interest. However as I downloaded your attached zip file, it only consisted of:
1. Cmlp_ad.mqh
2. SignalWZ_29.mqh
3. mlp_learn_r.mq5
and within mlp_learn_r.mq5, it requires the following files:
1. Expert.mqh
2. TrailingNone.mqh
3. MoneyFixedMargin.mqh
may I know how we can obtain them please?
Without them... the EA does not work.
Thank you! I am very grateful!
Dear Stephen,
Thank you for sharing your knowledge and work regarding this trading system!
I followed your article with great interest. However as I downloaded your attached zip file, it only consisted of:
1. Cmlp_ad.mqh
2. SignalWZ_29.mqh
3. mlp_learn_r.mq5
and within mlp_learn_r.mq5, it requires the following files:
1. Expert.mqh
2. TrailingNone.mqh
3. MoneyFixedMargin.mqh
may I know how we can obtain them please?
Without them... the EA does not work.
Thank you! I am very grateful!
they already exists under MQL include folder and you should add a header
#include <Expert\Expert.mqh> #include <Expert\Trailing\Expert.mqh> #include <Expert\Money\MoneyFixedMargin.mqh>
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
You agree to website policy and terms of use
Check out the new article: MQL5 Wizard Techniques you should know (Part 29): Continuation on Learning Rates with MLPs.
We wrap up our look at learning rate sensitivity to the performance of Expert Advisors by primarily examining the Adaptive Learning Rates. These learning rates aim to be customized for each parameter in a layer during the training process and so we assess potential benefits vs the expected performance toll.
We revisit and conclude our look at the role different formats of learning rates have on Expert Advisor performance by examining the adaptive learning rates and the one cycle learning rate. The format for this article will follow the approach we had in the last article by having test reports at each learning rate format section rather than at the end of the article.
Before we jump in, we’d like to mention some of the other critical machine learning design considerations that can hugely sway the performance of a model. One of these is the batch normalization of the input data. I had touched on this in prior articles as to why it's significant, however our next article will squarely dwell on this. For now, though, when designing the mode and format of a network, batch normalization is considered in tandem with the activation algorithms to be employed by the model or network. Thus far we have been using soft plus activation which has the tendency of producing unbound results i.e. unlike the TANH or Sigmoid activations that produce outputs in the ranges -1.0 to +1.0 and 0.0 to 1.0 respectively, soft plus can quite often produce results that do not pass the valid test and thus render the training and forecasting process invalid.
Author: Stephen Njuki