After some critical thoughts, I write about an article that shows me how machine learning technology is actually doing something very good for the world: Deep Learning helps protect honeybees. It’s about a topic that has been on everyone’s lips for a long time in IT: Automation!
Here the counting and evaluation of the honeycombs of the honey bees is automated and thus beekeepers get much easier information about the state of health of their hives. I think it’s so cool and I’m happy that researchers are working on it.
More and more often I read and hear that collecting data and setting up the data model are the biggest hurdles when using machine learning. It’s the same here.
The algorithms and tools are described, but not a word about infrastructure. That would interest me: Where do the tools run and what can be done better? I suppose that this case is well taken care of in a public cloud.
Maybe I will meet the author of this blog post at the Machine Learning Conference in Berlin.