How to get started with Machine Learning


Here are resources that are helping me get started with machine learning, and a few that I would have loved to have known about earlier. I'll probably be updating this page with new resources from time to time.

Stanford Cheat Sheets

A summary of terms, algorithms, and equations. (I barely understand the equations.=) These sheets, developed by Afshine and Shervine Amidi, differentiate between artificial intelligence (AI), machine learning (ML), and deep learning (DL) but many concepts overlap with each other. See this Venn diagram.

Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow: Concepts, Tools, and Techniques to Build Intelligent Systems

I highly recommend this book I'm going through at the moment, written by an ex-Googler who worked in YouTube's video-classification algorithm. It's dense but it introduces you to all relevant artificial intelligence, machine learning, and deep learning concepts, and guides you through preparing custom datasets to train algorithms, a bit of data science I guess. At the same time, it introduces you to three of the most-used machine learning frameworks—Sci-Kit Learn, Keras, and TensorFlow, being this last one the one I use on my day-to-day job developing and releasing machine learning models for production. Similar frameworks are Caffe or PyTorch, this one being used by Facebook developers. (Thanks to Keith Alfaro for the recommendation.)

Open-source code and tutorials

I got started with machine learning by trying open-source algorithms. It's common to visit the GitHub repository corresponding to a paper and give it a try. Two examples are Pix2Pix (2016) and EfficientDet (2020). You try to use their code as is, then try to use a custom dataset for training and see how the model performs for your needs.

TensorFlow re-writes many of these models and makes easy-to-follow tutorials.

  • Pix2Pix in TensorFlow Core - Made by the Google TensorFlow team, this tutorial offers you to View the code on GitHubDownload the Jupyter Notebook (written in Python) or Run the Notebook in Google Colab (where you can press a button in the cloud and see how each piece of Python code runs to understand the different parts of setting up and training an algorithm. Reading the dataset, peparing the training and validation set, creating the model, training it, and more).
  • TensorFlow tutorials - This is a good place to get your hands dirty. While machine learning has a strong theoretical component you can leave that aside and start by training and testing models for image classification, object detection, semantic image segmentation, and a lot more tasks.

Friendly user interfaces

  • Runway - A friend of mine, Cristóbal Valenzuela, is building his own machine learning platform for creatives. It's the place for people who don't know how to code (or don't want to) to be able to use complex machine learning models, training them with custom data and deploying them to the cloud. Here's an interview where he told me about the beginnings of Runway.
  • Machine Learning for Designers Talk - A talk I gave talking about these types of interfaces, a few projects, and the role they play for designers and people who don't know how to code.


Tutorials & live streams

  • Machine Learning Series YouTube playlist. Here is a compilation of some of the machine-intelligence-related video tutorials I've recorded.
  • Live Streams YouTube playlist. Weekly hands-on coding sessions on creative-coding, machine learning, art, design, and much more. From conceptual overviews to hands-on neural network architecture, automation, training, or cloud deployment.

Other resources

  • TensorFlow: Tensor and Image Basics - A video with basic tensor and image operations in TensorFlow. How to use tensors to encode images and matrices and visualize them.
  • TensorFlow: Visualizing Convolutions - A video to visualize the filters of an image convolution, an operation known for its ability to extract image features in an unsupervised way to perform classification tasks used in convolutional neural networks.
  • Awesome Machine Learning - A big and frequently-updated list of machine learning resources.
  • Suggestive Drawing - This is my Harvard's masters thesis, in which I explore how the collaboration between human and artificial intelligences can enhance the design process.

Found this post useful?

CodeMachine Learning