Federated learning, or what we can refer to as “collaborative learning,” is a method that uses local data stored on different servers to train an algorithm without exchanging the data. This method differs from traditional machine learning approaches, where all local datasets are uploaded to a single server.
In order to handle important challenges like data privacy, data security, data access rights, and access to heterogeneous data, federated learning enables different players to develop a single, strong machine learning model without sharing data.
What makes federated learning essential?
Traditional machine learning trains its algorithms using all of the training data that is stored on a central server. It functions well but has a number of shortcomings; for example, users expect quick responses, but the device and central server may have delayed connectivity. And, There is also the possibility of personal data breaches.
The application of federated learning
Usually, there are three industries using federated earnings: Fintech, healthcare, and IoT. Healthcare using federated learning in terms of critical information becomes more challenging with strict regulations.While Federated learning in IoT is to ensure IoT security and privacy.
Federated learning in Blockchain
Blockchain is a decentralised storage system that runs without the help of any centralised authority and keeps data in the form of a list of blocks connected by the cryptographic hash of the block before it. Data is kept in the blocks that make up the blockchain’s immutable chain. Every new transaction in the blockchain is added to the ledger and broadcast to all other network peers. Also, Bitcoin is a digital money that operates on the blockchain, a public, decentralised, and unchangeable ledger.
As I mentioned that safety-critical systems, including smart medical and healthcare systems, industrial environments, and smart cities, can achieve high potential privacy and security using the federated learning technique.Therefore, combining blockchain technology’s ability and ferderated learning can enhance security and privacy standards with minimum response time and cost.
Simple explanation of federated learning in this video
References :
Li, D., Luo, Z., & Cao, B. (2022). Blockchain-based federated learning methodologies in smart environments. Cluster computing, 25(4), 2585–2599. https://doi.org/10.1007/s10586-021-03424-y
