Machine learning algorithms are being developed and improved at an incredible rate but are not necessarily accessible to the broader community. That’s about to change thanks to DanKu, a new blockchain-based protocol for evaluating and purchasing ML models on a public blockchain such as Ethereum. A. Besir Kurtulmus explains how DanKu provides everyone access to high-quality, objectively measured machine learning models and has been used to launch the first fully trustless ML competition on the blockchain.
The DanKu protocol utilizes blockchain technology via smart contracts to create a decentralized and trustless marketplace for exchanging ML models. Anyone with a dataset, including software agents, can create DanKu contracts. The contract allows anyone to post a dataset, an evaluation function, and a monetary reward given to the best trained machine learning model for the data. Participants train deep neural networks to model the data and submit their trained networks to the blockchain. The blockchain executes these neural network models to evaluate submissions and ensure that payment goes to the best model. This gives ML practitioners an opportunity to monetize their skills directly and allows any participant or organization to solicit machine learning models from all over the world, incentivizing the creation of better machine learning models and making AI more accessible to companies and software agents.
Besir Kurtulmus is an algorithm engineer at Algorithmia helping making complicated things simpler by developing machine learning algorithms that are designed to scale. He also helps maintain the development environment for running machine learning models in the Algorithmia marketplace. Besir believes that machine intelligence will have a huge impact on our lives in the days to come and hopes to have a defining role in shaping this new future.
©2018, O'Reilly Media, Inc. • (800) 889-8969 or (707) 827-7019 • Monday-Friday 7:30am-5pm PT • All trademarks and registered trademarks appearing on oreilly.com are the property of their respective owners. • email@example.com