A quick reference for keeping track of the numerous ML algorithms out there! For each entry, the year is a rough indicator of when the term became relevant for the field of Machine Learning and is used for chronological ordering. This document will be continually updated.

Augmented Random Search (ARS)

2018 - Reinforcement learning using perceptrons instead of deep neural networks. For some specific applications, it is faster, with higher rewards. One example of an application for ARS is training robots to walk.

• Augmented Random Search —One of the Best RL Algs + What I Built (HackerNoon)
• Augmented Random Search Tutorial - How to Train Robots to Walk! (YouTube)

Generative Adversarial Networks (GAN)

2014 - Used primarily in the field of computer vision, GANs learn based on two adversarial neural networks, a generator and a discriminator. At its most fundamental level, the generator tries to “fool” the discriminator while the discriminator tries to “catch” the generators “lies.” Learning is reinforced through feedback loops.

• A Beginner’s Guide to Generative Adversarial Networks (GANs)
• An intuitive introduction to Generative Adversarial Networks (GANs)

Long Short-Term Memory (LSTM)

1997 - A type of Recurrent Neural Network (RNN) used in supervised deep learning, makes predictions based on time-series data, essentially learning over time base on prior knowledge. This kind of ML is used in Siri and Amazon Alexa.

• A Beginner’s Guide to LSTMs and Recurrent Neural Networks
• The magic of LSTM neural networks

Convolutional Neural Networks (CNN)

1989 - A class of deep neural networks used for image processing, including object classification and facial recognition

• A Beginner’s Guide to Convolutional Neural Networks (CNNs)
• An Intuitive Explanation of Convolutional Neural Networks
• What are Convolutional Neural Networks?