Machine Learning is often seen as the exclusive preserve of Data Scientists. Danilo Poccia disagrees. To learn the basics he will be speaking at Voxxed Days Zurich on the 23rd of February. We asked him what makes him think non-mathematicians and scientists can get started with Machine Learning.
Do you think machine learning is accessible to everyone? Do I need a PhD? How easy is it to get started?
Machine Learning – or how we can make a computer program find patterns in data, and learn – is a broad and interesting topic. There is a lot of mathematics and statistics in the algorithms that you can use. However we have reached a point where tools and services are so easy to use that you don’t need to understand all the implementation details. This is in the same way as a good car mechanic doesn’t know all the physics of a combustion engine. For example, you can follow a quick tutorial and start using open source tools such as Apache MXNet to do image recognition.
What excites you the most in the machine learning field?
It is such a fast growing field. Things that were considered almost impossible a few years ago are now getting mainstream. For example, speech recognition with platforms such as Amazon Alexa is getting to a point where people are comfortable to use conversational interfaces instead of smartphones or keyboards, and interact with technology in a more intuitive way. Most of those advancements are possible thanks to something called “deep learning”. This is a branch of machine learning that made it possible to train and use more complex (deeper) neural networks to model high level abstractions in data. This can then apply that to multiple fields such as healthcare, or the car industry.
How did you get started with machine learning?
I was always passionate about Artificial Intelligence. I started reading classical books such as “Artificial Intelligence: A Modern Approach” by Stuart Russell and Peter Norvig, or the works of Marvin Minksy. When the first AI and ML massive online courses were available in 2011, I did both. It was so exciting to get training from people such as Peter Norvig, Sebastian Thrun, and Andrew Ng!
What can attendees hope to learn at your session?
I am going to help developers who have no previous knowledge in the field. Together we can understand the terminology and the normal workflow of machine learning practitioners. We will have a specific focus on the new deep learning techniques. You don’t need to understand all the mathematics and statistics used in this field to build amazing things. We’ll see together how open source platforms, such as Apache MXNet, and cloud services, such as those in the Amazon AI family, bring natural language understanding, automatic speech recognition, visual search and image recognition, text-to-speech, and more, within reach of every developer. Let’s use all those tools to build smarter apps.
For an introduction to Machine Learning, join us at Voxxed Days Zurich.