drawing

The simplest explanation of machine learning you’ll ever read was first published May 24, 2018 on HackerNoon

Author

Cassie Kozyrkov

“The simplest explanation of machine learning you’ll ever read” is one of many articles published on HackerNoon by author Cassie Kozyrkov. According to her Twitter account, Dr. Kozyrkov is the Head of Decision Intelligence at Google. She also founded this division! She is based in New York but was born in Russia and grew up in South Africa. According to her LinkedIn profile, she has studied at U Chicago, NC State, and Duke. Her PhD is in Psychology and Neuroscience from Duke.
Cassie has an extensive online presence and has published articles, tutorials, etc on many platforms. She is active on social media like Twitter, but also writes for sites such as HackerNoon or Towards Data Science.

Summary

What is machine learning? Is it the same as Artificial Intelligence? Not quite. Machine learning and artificial intelligence might sound scary, but they don’t have to be. As Cassie Korzkykov says, all machine learning is is a “thing-labeler”. That is, you give an algorithm some information, and it spits out how you should label it. Artificial Intelligence falls under the umbrella of Machine Learning.
Developing Machine Learning and AI techniques means that we are able to get labels with humans putting in less and less work. Rather than giving instruction through code, machine learning and AI operate on examples. These examples give patterns for which machine learning develops an algorithm.
To demonstrate this idea, the author gives a photo of a cat, and notes that with machine learning and AI, you’d be able to give a bunch of photos with cats and a bunch without. Machine learning could take those training photos, and be able to form a sort of formula that it could use for any other photos going forward.

Random plot using R’s mtcars dataset

mtcars using ggplot2

Interactive data table using mtcars

DT::datatable(mtcars)