Metatrader 5 / Metatrader 4 for MQL5 / MQL4 articles preview

The concepts of trend and flat can be clearly and fully defined. Besides, it is possible to compare the trendiness degree of different instruments as well as of a single instrument on different scales. Such an analysis allows us to evaluate characteristics and features of instruments and construct trading algorithms considering individual behavior of each instrument.

Knowing statistical characteristics of an instrument and their change over time, we can get rid of optimizing parameters when developing a trading algorithm. Instead, we can perform measurements and deliberately change trading algorithm parameters depending on statistical characteristics of a traded instrument.

more...

The MetaTrader 5 features the ability to create custom symbols having their own quotes and ticks. They can be accessed both from the terminal interface and at the program level, via MQL API. Custom symbols are displayed on standard charts, allowing the application of indicators, marking with objects, and even creation of trading strategies based on these symbols.

Custom symbols can use real symbol quotes provided by brokers or external resources as data sources. In this article, we will consider several popular methods for transforming working symbols that provide additional analytical tools for traders:

• equal volume and equal range chart
• tick charts
• time shift of quotes with the conversion of candlestick shapes
• Renko

Additionally, we will develop a mechanism for adapting Expert Advisors to trading a real symbol which is associated with a derivative custom symbol on whose chart the EA is running.
In this article, source (standard) symbol charts use a black background, and custom symbol charts use a white background.

more...

... Therefore, let us consider how to adhere to a scientific approach and avoid superstitions during the development of a trading algorithm. We will consider some ideas in this article using a simple trading system as an example.

more...

I am a developer of automatic strategies and software with over 5 years of experience. In this article, I will open the veil of secrecy to those just starting to trade in Forex or on any other exchange. Besides, I will try to answer the most popular trading questions.

I hope, the article will be useful for both beginners and experienced traders. Also, note that this is purely my vision based on actual experience and research.

more...

Contents

• Introduction
• 1. Distinctive features of convolutional neural networks
  
  • 1.1. Convolutional layer
  • 1.2. Subsampling layers
  
  
• 2. Principles of training neurons in convolutional layers
• 3. Building a convolutional neural network
  
  • 3.1. Base class of neurons
    
    • 3.1.1. Feed-forward
    • 3.1.2. Error gradient calculation
    
    
  • 3.2. Subsampling layer element
    
    • 3.2.1. Feed-forward
    • 3.2.2. Error gradient calculation
    
    
  • 3.3. Convolutional layer element
  • 3.4. Creating a convolutional neural network class
    
    • 3.4.1. Convolutional neural network class constructor
    • 3.4.2. Convolutional neural network forward propagation method
    • 3.4.3. Convolutional neural network backward propagation method
    
    
  
  
• 4. Testing
• Conclusion
• List of references
• Programs used in the article

more...

----------------
Previous two articles related to this subject:

• Neural Networks Made Easy - MT5
• Neural networks made easy (Part 2): Network training and testing - MT5

In this article, we will search for market patterns, create Expert Advisors based on the identified patterns, and check how long these patterns remain valid, if they ever retain their validity. I think this article will be extremely useful for those who create self-adapting trading systems. I would like to start with an explanation of what I mean by the concept of brute force in the context of forex trading. In general, in the brute-force approach we want to determine some sequence of numbers for a code or any other purpose, which ultimately will allow to achieve the desired result with the maximum probability or with the maximum available probability, using this sequence. The desired result can be to get cryptocurrency or to hack an account or Wi-Fi password. The application area is very wide. In the case of forex, our sequence should give the maximum profit, while working as long as possible. The sequence can be of any type and length. The only required condition is its efficiency. The sequence type ultimately depends on our algorithm

more...

In this article, I am going to use math and programming to delve into these strategies and evaluate their profitability. The article features mathematical and practical sections. In the math section, I will provide equations for calculating the expected payoff of the strategies and other important parameters not taken into account by many traders. In the practical part, I will develop a simple grid and martingale, and compare the equations with reality. The article is especially useful for beginners since these strategies are usually the first ones they come across. Blind faith into them may lead to disappointments and waste of time, as happened to me back in the time. If I had not known math, perhaps I would still have believed in them. But these strategies still have a rationale if you look at them correctly. This is what I am going to prove.

more...

In the previous article entitled "Practical application of neural networks in trading. It's time to practice", we considered the practical application of a neural network module implemented using Matlab neural networks. However, that article did not cover questions related to the preparation of input data and network training related operations. In this article, we will consider these questions using examples and will implement further code using neural networks of libraries working with Python. This time, the trading system implementation principles will be different. This variant was briefly described in paragraph 5 of the basic article "Practical application of neural networks in trading". For this implementation, we will use the TensorFlow machine learning library developed by Google. We will also use the Keras library for describing neural networks.

more...

Contents

• Introduction
• 1. The problem
• 2. Experiment 1
• 3. Experiment 2
• 4. Experiment 3
• Conclusions
• References
• Programs used in the article

1. Neural networks made easy - MT5
2. Neural networks made easy (Part 2): Network training and testing - MT5
3. Neural networks made easy (Part 3): Convolutional networks - MT5
4. Neural networks made easy (Part 4): Recurrent networks - MT5
5. Neural networks made easy (Part 5): Multithreaded calculations in OpenCL - MT5
6. Neural networks made easy (Part 6): Experimenting with the neural network learning rate - MT5

more...

Contents

• Introduction
• 1. Distinctive features of adaptive optimization methods
  
  • 1.1. Adaptive Gradient Method (AdaGrad)
  • 1.2. RMSProp method
  • 1.3. Adadelta Method
  • 1.4. Adaptive Moment Estimation Method (Adam)
  
  
• 2. Implementation
  
  • 2.1. Building the OpenCL kernel
  • 2.2. Changes in the code of the main program's neuron class
  • 2.3. Changes in the code of class not using OpenCL
  • 2.4. Changes in the code of the main program's neural network class
  
  
• 3. Testing
• Conclusions
• References
• Programs used in the article

more...