SUPERALGOS GOVERNANCE
DJ05GaZ Proof of Value, May 2022
Bitcoin Factory Test Client & Server Testing
Bitcoin Factory Dedicated Hardware
As the bitcoin factory was flourishing into a live project it seemed necessary to dedicate hardware to this project, but after testing on a couple of home computers it dawned on DJ05GaZ that he needed something a little faster. A server was purchased to get the project off the ground. Currently this server has dual Xeon 12 Core CPU’s with 300GB RAM and is completely dedicated to this project.
Testing The Client
Initially Luis ran the test server and the ML team provided computing power to test cases provided by the server. A total of 4 rounds and over 8000 cases were successfully solved. During this period issues appeared regarding building test cases and providing them to the team efficiently.
Testing The Server
Luis coded extra features that enabled the test server and network node to be ran from team members and cloud computers. This was a huge step as it enabled the project to scale easily without bottle-necking. Tigers40 setup a network node from a cloud service that connects the Test Clients to the Test Server with high reliability. DJ05GaZ started running a Test Server with multiple nodes using the extensive data mining indicator list, created by theblockchainarborist.
Bumps in the Road
Testing in its infancy has proven quite a task for newcomers and coders alike, but it goes without saying that the team is steadily progressing. Members are finding and fixing problems more easily and are creating a well oiled system in the process. The whole team deserves a high five for the dedication and hard work they have put in to get this project running.
Data Mine Indicators
As already reported by theblockchainarborist, the most recent version of the machine learning project saw the introduction of testing indicators, which were taken from a shortlist of data mines found in Superalgos. Currently the system is processing over 1000 test cases with the Fibonacci indicators activated. A total of 34 data mines have been added to the indicator list.
Test cases are to be deployed to the clients with selected indicators turned on. This will produce results that can be analysed, allowing the system to develop models that achieve good results with high accuracy forecasting.
What Next?
Testing with indicators from a single server is not an ideal scenario, tracking test servers so that anyone running a test server can see which indicators have already been tested and which indicators need to be tested.
Report analysis for better forecasting.
Multiple test servers.
More test clients.
Integrated ML Dashboard.
The Superalgos Machine Learning Group
The machine learning project is open to everyone regardless of skills, if you can contribute hardware, good. If you can debug, even better but overall there’s a place for you in the machine learning group. Why not join the telegram group and see what this amazing project has to offer!
