Machine Learning

This video from Google I/O 2019 is an excellent 100 foot overview

What is Machine Learning?

In Traditional Programming you write rules and use data to provide answers: If the fingers in a hand are in a V shape, then that is scissors.

In Machine Learning you use answers and data to help a machine learn the rules: Hello machine, see this collection of images of hands? These are all "scissors" and any future images that share characteristics like these are also "scissors."

Phases

Machine Learning typically has two phases, which sometimes occur consistently with feedback.

Training Phase

Answers and Data are fed into the machine to infer the rules. This machine is now a model!

Inference Phase

The model will now make predictions from new data (with confidence).