I recently finished my first machine learning project! I was able to follow along with Robin Sloan’s writing with the machine project, and I trained a different dataset, to some hilarious results. It was a lot of fun, and now I’m trying to dive more into neural networks and figure out more projects I can try. If you’re interested, here are some resources and tools to learn more about machine learning, and the types of cool things you can do with it.
About Machine Learning
- “Shark or baseball? Inside the ‘black box’ of a neural network” on Wired
- “Artificial intelligence: the art world’s weird and wonderful new medium” on How to Spend It
Projects
- “Warner Music Group Signs an Algorithm to a Record Deal” on Rolling Stone
- “Teaching the Computer to Play the Chrome Dinosaur Game with TensorFlow.js Machine Learning Library” on InfoQ
- “Exploring Neural Networks with Activation Atlases” on Distill
- “Microsoft Word will change your words to be ‘gender inclusive’” on The Telegraph
- “How to Automate Tasks on GitHub With Machine Learning for Fun and Profit” on Towards Data Science
- “There’s a subreddit populated entirely by AI personifications of other subreddits” on The Verge
- “Oh no, Samsung’s AI lab can create a video of you from a single still photo” on Fast Company
Tools
- TensorFlow playground
- ImageNet (database of images you can use)
- Rectified Linear Unit
- Machine Learning Repository
- Wekinator (open source software for using machine learning to build musical instruments)
