MPM-4243 Blending Event Stream Processing with Machine Learning using the Kafka ecosystem | Voxxed Days

Voxxed Days Milano 2019
on Saturday 13 April

   Blending Event Stream Processing with Machine Learning using the Kafka ecosystem

Conference

Big Data & Machine Learning
Big Data & Machine Learning
Intermediate level
Aula 7 Saturday from 12:30 til 13:20

Kafka comes today as the essential landmark whenever the main task is to write an events stream application, to build a streaming architecture, to move software toward realtime processing. Machine learning is made of several topics: data pre-processing, training, validation, scoring, deployment and so on; given that some of these aspects easily fit in a streaming scenario many others do not because they have been traditionally grown surrounded by historical data and supported by batch architectures (training). After a brief introduction about "streamophobic" artificial intelligence critical topics, the talk will describe how the latter can be applied in realtime by naturally exploiting the Kafka architecture, detailing the proposed technical solutions about online machine learning training and dynamic model scoring with modern oss ML systems as H2O, justifying this effort by listing the outcoming value, as enabling Kappa Architecture and allowing users to manage end-to-end machine learning applications in a pure streaming context. The talk will conclude by presenting the current production use-cases hosting the targeted solutions.

Kafka   Machine learning   Data Streaming API  
Sign-in
Make sure to download the Android or iOS mobile schedule.