Dmitriy Gizlyk / Profil

We continue to study unsupervised learning algorithms. Some readers might have questions regarding the relevance of recent publications to the topic of neural networks. In this new article, we get back to studying neural networks.

We continue considering association rules. In the previous article, we have discussed theoretical aspect of this type of problem. In this article, I will show the implementation of the FP Growth method using MQL5. We will also test the implemented solution using real data.

As a continuation of this series of articles, let's consider another type of problems within unsupervised learning methods: mining association rules. This problem type was first used in retail, namely supermarkets, to analyze market baskets. In this article, we will talk about the applicability of such algorithms in trading.

In this part we continue discussing Artificial Intelligence models. Namely, we study unsupervised learning algorithms. We have already discussed one of the clustering algorithms. In this article, I am sharing a variant of solving problems related to dimensionality reduction.

In the previous article, we have created a class for data clustering. In this article, I want to share variants of the possible application of obtained results in solving practical trading tasks.

We continue to consider the clustering method. In this article, we will create a new CKmeans class to implement one of the most common k-means clustering methods. During tests, the model managed to identify about 500 patterns.

It has been more than a year since I published my last article. This is quite a lot time to revise ideas and to develop new approaches. In the new article, I would like to divert from the previously used supervised learning method. This time we will dip into unsupervised learning algorithms. In particular, we will consider one of the clustering algorithms—k-means.

In the previous article, we started considering methods aimed at improving neural network training quality. In this article, we will continue this topic and will consider another approach — batch data normalization.

As the next step in studying neural networks, I suggest considering the methods of increasing convergence during neural network training. There are several such methods. In this article we will consider one of them entitled Dropout.

Perhaps one of the most advanced models among currently existing language neural networks is GPT-3, the maximal variant of which contains 175 billion parameters. Of course, we are not going to create such a monster on our home PCs. However, we can view which architectural solutions can be used in our work and how we can benefit from them.

We have previously considered the mechanism of self-attention in neural networks. In practice, modern neural network architectures use several parallel self-attention threads to find various dependencies between the elements of a sequence. Let us consider the implementation of such an approach and evaluate its impact on the overall network performance.

We have already passed a long way and the code in our library is becoming bigger and bigger. This makes it difficult to keep track of all connections and dependencies. Therefore, I suggest creating documentation for the earlier created code and to keep it updating with each new step. Properly prepared documentation will help us see the integrity of our work.

In previous articles, we have already tested various options for organizing neural networks. We also considered convolutional networks borrowed from image processing algorithms. In this article, I suggest considering Attention Mechanisms, the appearance of which gave impetus to the development of language models.

In previous articles, we used stochastic gradient descent to train a neural network using the same learning rate for all neurons within the network. In this article, I propose to look towards adaptive learning methods which enable changing of the learning rate for each neuron. We will also consider the pros and cons of this approach.

We have previously considered various types of neural networks along with their implementations. In all cases, the neural networks were trained using the gradient decent method, for which we need to choose a learning rate. In this article, I want to show the importance of a correctly selected rate and its impact on the neural network training, using examples.

We have earlier discussed some types of neural network implementations. In the considered networks, the same operations are repeated for each neuron. A logical further step is to utilize multithreaded computing capabilities provided by modern technology in an effort to speed up the neural network learning process. One of the possible implementations is described in this article.

We continue studying the world of neural networks. In this article, we will consider another type of neural networks, recurrent networks. This type is proposed for use with time series, which are represented in the MetaTrader 5 trading platform by price charts.