From the course: Advanced Data Processing: Batch, Real-Time, and Cloud Architectures for AI
Join today to access over 24,900 courses taught by industry experts.
AI using hybrid computing
From the course: Advanced Data Processing: Batch, Real-Time, and Cloud Architectures for AI
AI using hybrid computing
- [Instructor] How does hybrid computing work for different facets of AI ML? We can use hybrid computing to train and serve ML models across different types of environments like cloud, on-premise and edge. Such architectures should be built to ensure reliability and efficiency. How does feature engineering work in a hybrid setup? Feature engineering may use a distributed set of data sources, and these data sources can be from devices, on-premise databases, or cloud services. Feature engineering can itself be distributed. For example, each device that has the data source can execute feature engineering tasks locally before shipping the results to a central server. A central server can aggregate data from multiple devices and environments to create the feature store When processing data in multiple environments, it is important to ensure that the data is protected during processing, storage, and transfer. Data privacy should be ensured at all stages of feature engineering, especially…
Download courses and learn on the go
Watch courses on your mobile device without an internet connection. Download courses using your iOS or Android LinkedIn Learning app.
