The main technical progress of the DeepBrain Chain. Successfully testing a machine learning model in real life
DeepBrainChain project talks about its developments, prospects and events.
Artificial Intelligence — the result of processing data arrays using algorithms. Computing power and algorithms are not only essential in achieving responsiveness and materialization of relevant functions in AI, but are fundamental building blocks. Currently, the high price of equipment severely limits AI companies, often an insurmountable threshold for startups..
The AI industry needs solutions that are efficient and save money. Deepbrain chain — the world’s first computing platform for artificial intelligence powered by blockchain technology. our company offers a solution for reliable and uninterrupted computing power for AI and the companies that work with it. We connect tens of thousands of miners around the world to create a single powerful computing network.
On June 3, DeepBrain Chain (DBC) successfully ran three real-life machine learning trials on its testnet.
we conducted tests for learning artificial intelligence in the real world
In the first test, our artificial intelligence was able to successfully recognize images from the voluminous MNIST handwriting sample database. The MNIST dataset comes from the State National Institute of Standards and Technology (NIST). The teaching set is composed of 250 different handwritten numbers, of which 50% are written by high school students, and another 50% — by the staff of the United States Census Bureau. The handwritten numbers data in the test set have the same ratio.
Next, we launched an NLP model for classifying Chinese text. He performed a multitasking solution using a convolutional neural network (CNN) as a learning model.
In test mode, AI DeepBrain Chain successfully launched a ready-made doc2vector model for document embedding. By entering a text document, the model exports the document matrix vector to perform testing and use the trained model as a target.
Our artificial intelligence will start working at the end of June
Despite the obvious complexity of the project, based on the work that we have done while developing and developing our project, we can say with confidence that our project was launched on time.
We have completed the first round of testing the iteration. We solved all the issues that arose as a result of testing.
Networking functions for P2P have been identified, and a second round of testing and testing of additional P2P network functions has begun. In addition, a second iteration was introduced and the construction of a structural design began based on the needs discussed in blockchain and AI community.
Here are the events we have prepared for June
The list of wallets for buying rights to AI Mining Machines is introduced on June 11.
Test network machine learning — end of June.
Consensus node selection kicks off mid-July.
Mainnet launch slated for late October.
Backpropagation and Deep Learning in the Brain
In addition, DeepBrain Chain will help 1–2 promising AI projects to eventually build a global ecosystem.
Who we are?
we — the world’s first AI computing platform built on blockchain technology. We use blockchain technologies, help companies save up to 70% of computing costs while protecting confidential data for machine learning. The main idea of the project is to create «Decentralized cloud platform for AI computing» and «AWS to AI».
DeepBrain Chain aims to become the mainstream and publicly available network for delivering high performance computing and protecting the confidential data of AI companies and all AI users..
AI companies will be able to use their own tokens on top of the DeepBrain Chain and offer their services and AI products.
The income of mining nodes in the DeepBrain Chain ecosystem is in two parts — system reward and consumer-paid DBC for AI computing power.
AI — Artificial Intelligence.
Huobi.pro, Kucoin, LBank, gate.io, Allcoin, Bitbns, Switcheo, Abucoins
NEO Tracker, O3, NEO-GUI, NEON, Hyper Pay, ROOTOKEN, Morpheus