Klassisches Machine Learning funktioniert am besten mit strukturierten Daten. In heutigen Analyseszenarien nimmt aber der Anteil unstrukturierter Daten immer stärker zu. Das Plugin Machine Intelligence kann zusammen mit Pentaho Data Integration Deep Learning-Operationen auf Daten aller Art ausführen. Auf dem Pentaho User Meeting am 26. März stellt Ken Wood das Plugin vor.
Ken, who are you?
I am the head of Hitachi Vantara Labs (formerly, Pentaho Labs) where we do research, development and innovation (Rd&I) for all things Pentaho and Hitachi related. Most recently, we have been focused on expanding the use cases that Pentaho Data Integration (PDI) can address by adding machine learning natively to PDI. We call this capability Plugin Machine Intelligence (PMI).
Since PMI’s release last year, we have been continuing to expand its capabilities and looking for new and innovative ways to bring more data science capabilities to Pentaho. We call it „machine intelligence“ because machine intelligence is more than just training and testing models. Machine intelligence is about solving real problems and providing real insight from data, or what as I like to say, „…putting the models to work!“
What connects you to Pentaho?
Fun fact about me, in late 2013, I started doing the technical evaluation of Pentaho for acquisition by Hitachi (which happened in early 2015). I have been working for Hitachi for almost 22 years, in several technical & strategic roles in infrastructure, analytics, IoT and AI, and in 2016, Pentaho asked me to run Pentaho Labs.
Which topic will you cover in your talk?
My talk at Pentaho User Meeting will be to introduce you to PMI and discuss our motivation behind creating this powerful tool, and show a demonstration or two of how to put machine learning to work to build applications. Finally, I’ll share with you how we envision PMI and Pentaho, from a Hitachi Vantara Labs perspective, may evolve in the future.