Conference about Big Data, High Load, Data Science, Machine Learning & AI
Juan Tomas Garcia is a Chief Envisioner at
Deepmind: Alpha Go to Alpha Zero
Alpha Go is possibly one of the most important AIs that have been developed lately. Its development and the challenge of winning the 18-time world champion Lee Sedol is so epic that it is worth mentioning what is Alpha Go, how it works and how it has evolved to Alpha Zero.
Alpha Go is a reinforcement learning model that uses convolutional networks and Monte Carlo Tree Search. It is one of the simplest and most elegant models that have been developed to solve a problem as complex as the GO game.
In the end, we will describe how he has evolved to become Alpha Zero and become a champion of other games.