|
PROJECT
SUPERVISORS
| EECE 597 | Prof. Z. Wang - ECE | ||
| To Apply: | For information & availability of specific projects. | ||
| ID | Status | Name | |
| ZW-1 | Available | Scalable Federated ML by Leveraging Blockchain Technology | |
| Comparing with the existing machine learning system where raw data is saved and processed in the central servers, the federated machine learning has been proposed to process the raw data close to data sources in a distributed manner and then merge the learning results together. While, it is not very scalable when considering the tremendous Internet of Things (IoT) devices with the requirements of data privacy and integrity in near future. On the other hand, the blockchain and smart contract technology are featured by tamper-proof and able to run the same piece of bytecode on decentralized virtual machines, so they could be applied to promote the federated machine learning to a much larger scale. While, when merging the learning result together in a large scale, the read-write conflict may happen. Improperly solving this problem may result in either very slow learning speed or mistakenly overwriting the learning results contributed by others. Students are expected to finish the EECE571B and EECE571G courses to learning the fundamental knowledge and software engineering skills regarding blockchain, then work with the PhD students in the lab to address the problem by designing an application protocol to coordinate the knowledge-contributing devices so that the decentralized machining is scalable and reliable. | |||
