In this article, we will talk about how to easily and quickly find and use patterns in huge amounts of statistical data using the "Search for patterns in statistical data" technology.
Today, a lot of statistical data has accumulated, in which you need to find patterns. This is the subject of this article. The technology we are talking about is called "finding patterns in statistical data". Some aspects of this technology overlap with neural networks, such as preparing data for processing.
You can get acquainted with the operation of this technology by downloading the program from the link provided at the end of the article.
The video describes in detail how this technology works using a specific example. The process is demonstrated from start to finish, with thorough explanations.
As an illustration, the task of finding patterns in statistical data on exchange trading for 16 years was chosen.
Now let's discuss the technology itself. The solution to this problem can be divided into three stages.
- Data collection.
- Data analysis and processing, identification of patterns-algorithms.
- Application of ready-made algorithms for solving specific problems.
Data collection
We have statistical data that must be prepared before processing. This technology works in a binary system, that is, it uses the principle of "yes" or "no", " 0 " or "1". To do this, you need to set various conditions in the data and determine whether they are met or not. If the condition is met, the value is 1. If not, it is 0. The video shows in detail how data is prepared for subsequent processing. As a result, we get a table ready for processing.
Data analysis and processing, identification of patterns-algorithms
Then we adjust the parameters and process the table to identify patterns, which are further called algorithms. The video shows how you can analyze each algorithm and understand how it works. Depending on the number of statistics and settings that we have set, we can get thousands of algorithms.
Application of ready-made algorithms for solving specific problems
There are many ways to use ready-made algorithms.
You can use a single algorithm, or create a package of several algorithms, and apply all the found algorithms at once.
In the video, you will see all the available options for using algorithms.
There are several approaches to using algorithms.
Generating algorithm packages
You can create working algorithm packages based on all the algorithms found. Such packages will work effectively with a given probability in the statistical data under study, as it was shown in the previous video.
The video explains in detail how such packets are formed.
Using the full draft algorithm package
This method is similar to the work of neural networks: the input is given parameters of the object, and at the output we get the probability of the object's truth.
This approach uses nonlinear regression with a natural logarithm. After processing the entire set of algorithms, we get the formula:
Y = A + B * ln(X),
where the coefficients A and B are constant for a given set of algorithms,
and X is the number of algorithms that will work in this object.
By substituting the values A, B, and X into this formula, we can determine the probability of how this object will perform in statistical data.
This approach is based on the following principles:
1. Each object can run a different number of algorithms.
2. The more algorithms worked in the object, the higher the probability of a positive result.
The video explains in detail how this approach works.
Generating packets from chunks
This approach is based on the following principle.
If you have a large amount of statistical data, you can divide it into small fragments and search for algorithms in them. Then you can test each algorithm on all statistical data and select only those that show good results.
The video explains in detail how this approach works.
Analysis of algorithms
Depending on the settings that we set, you can get a lot of different algorithms.
This video analyzes how and what different settings affect when searching for algorithms.
Use this link to download the program to evaluate the capabilities of this technology.
https://ru.files.fm/filebrowser#/Search%20for%20patterns
Top comments (0)