Carsten FriedrichPart 8 — Tic Tac Toe with Policy Gradient DescentDo you really need values if you have good policies?Jul 20, 2018Jul 20, 2018
Carsten FriedrichPart 7 - This is deep. In a convoluted way.One trick that we haven't tried yet is the use of Convolutional Neural Network (CNN) layers. Time to do so now.Jun 6, 2018Jun 6, 2018
Carsten FriedrichPart 6 - Double Duelling Q Network with Experience ReplayIn the previous part we discovered that just being a bit less greedy in our action policy was not enough.Jun 6, 2018Jun 6, 2018
Carsten FriedrichPart 5 - Q Network review and becoming less greedyLooking at what went wrong and becoming less greedyJun 6, 2018Jun 6, 2018
Carsten FriedrichPart 4 - Neural Network Q Learning, a Tic Tac Toe player that learns - kind ofTraining a Neural Network to learn the Tic Tac Toe Q functionJun 6, 20182Jun 6, 20182
Carsten FriedrichPart 3 - Tabular Q Learning, a Tic Tac Toe player that gets better and betterIn this part, we will introduce our first player which actually uses a machine learning approach to playing Tic Tac Toe.Jun 6, 20183Jun 6, 20183
Carsten FriedrichPart 2 - The Min Max AlgorithmIn this Notebook, we will introduce and then use the Min-Max algorithm to create a computer player which will be able to play Tic Tac Toe.Jun 6, 2018Jun 6, 2018
Carsten FriedrichPart 1 - Computer Tic Tac Toe BasicsBasic Tic Tac Toe support classes and game logicJun 6, 20181Jun 6, 20181
Carsten FriedrichTeaching a computer to play Tic Tac ToeFrom classic algorithms to Reinforcement learning with Neural NetworksJun 6, 20181Jun 6, 20181