TensorFlow and Deep Reinforcement Learning, Without a PhD - Google I/O '18

Dive into a 41-minute conference talk from Google I/O '18 exploring TensorFlow and deep reinforcement learning. Learn how to teach a neural network to play the video game Pong using only pixel data, without prior rules or strategy coaching. Discover the cutting-edge technique of reinforcement learning, which bridges academic deep learning and natural learning processes. Explore topics including dense neural networks, activation functions, policy gradients, gradient descent, and training loops. Gain insights into the practical application of machine learning tools like Cloud Machine Learning Engine. No PhD required to understand this accessible introduction to complex concepts at the forefront of deep learning research.

Intro
Dense neural network
Activation functions
Success?
Pong?
Policy gradients
Policy network
Gradient descent
Playing a game
Training loop
Learned weights
Tools: Cloud Machine Learning Engine