Anyone that creates neural nets in probably aware of (and uses) Tensorflow. It’s Google’s open source library for machine learning – it eases the creation and calculation of models using CPUs and GPUs. It really is quite brilliant and is a must learn.
However, today let’s take a step back and assume you don’t know what the heck a neural net is (kudos for making it this far if you didn’t understand what was going on in the first paragraph). Neural nets are part of the machinery that is driving this big AI boost you may have heard of. In the form I’ll discuss it’s really just a way to take inputs X_1 and X_2 and map that to some output Blue or Orange. For example, we could (in theory) take historical trading data (X_1 is trading volume yesterday, X_2 is stock return yesterday) and get a prediction for whether a stock will go down (Blue) and up (Orange).
There are many, many ways to do this and I don’t want to get into all of that. What I think you should see is this little sandbox put together by the people at Tensorflow so you can see how different combinations of features, hidden layers, and neurons can create predictions that are scarily precise. Don’t know what those things are? No problem! Just go to the playground and click around – I am fairly certain you’ll be able to figure out what is going on.
http://playground.tensorflow.org/