javaBin Online: "Introduction to Deeplearning4j" with Marek Będkowski

You are hereby invited to our eighth javaBin Online session with our guest Marek Będkowski!

All javaBin Online sessions are streamed over YouTube, so you can enjoy them from the comfort of your home office. Just before we start we'll post the YouTube link in the comments for this Meetup. We'll also set up https://www.java.no/online to redirect to the talk.

AGENDA

• 18:00: Welcome from javaBin
• 18:05: "Introduction to Deeplearning4j" with Marek Będkowski
• 19.00: Q&A

ABSTRACT
AI and deeplearning have steep learning curve due to domain complexity and domination of Python-like approaches. For one of my projects we decided to skip one part of this equation and try if we can come up with working solution using deeplearning4j. This proved to be good choice as we could focus on deep learning itself. During this presentation I'd like to encourage you to do your own experiments by proving that some of this stuff is not really needed in the early stages of deep learning journey and that using dl4j can allow you to get you started with known tooling.

SPEAKER BIO

I've been developing software for over 15 years and over 10 years within Java Ecosystem. At some point I developed a Live Streaming solution from scratch which turned into AI Research project with DeepLearning4j as it's deep learning framework. I spent countless hours reading papers and trying to adjust some of sota networks to my problem. Ever since I don't believe in AI taking control over the world soon :) Secondly I always try that despite its computational complexity deep learning is hard and that in the beginning one doesn't need to know all theory related to it.