Leveraging Machine Learning
ML applications have grown to dominate in highly visible and enterprise-scale uses today: Google search results, Facebook/Instagram/TikTok/Twitter sorting algorithms, YouTube’s suggested content, Alexa/Siri voice assistants, internet advertising, and more. These use cases all host their ML models and perform their inferences in the cloud, then show the results to us as end users on our edge devices such as phones, tablets, or smart speakers. This paradigm is beginning to change, with more models being stored (and inferences being run) on the edge devices themselves. The shift to processing at the edge removes the need for transmission to and storage in the cloud, and as a result, provides the benefits listed here:
- Enhanced security: Reducing attack vectors
- Enhanced privacy: Reducing the sharing of data
- Enhanced performance: Reducing application latency
This chapter is intended to give an overview and practical guide for using...