Seminář Laboratoře softwarových architektur a informačních systémů
Week 4 - Architectures of Real-World Machine Learning Systems in the Cloud (Lukáš Grolig)
Abstract. Applied Machine Learning (ML) is a different field than writing another neural network algorithm. One algorithm is never enough to handle all requirements. A system is usually created by connecting multiple existing systems and multiple ML algorithms. The running system in the cloud brings another level of complexity. Some things can be used out of the box, things that have to be written and a lot of optimizations done so running costs are kept within a reasonable range.
In this talk will be discussed the most important architecture concerns building real-world ML systems.
Speaker: Lukáš Grolig, entrepreneur, software architect, data scientist and Ph.D. student
PV226-03-Grolig.pdf