Breaking News
Christmas Truce of 1914, World War I - For Sharing, For Peace
The State's Last Stand
The Roots of Collectivist Thinking
What Would Happen if a Major Bank Collapsed Tomorrow?
Top Tech News
Travel gadget promises to dry and iron your clothes – totally hands-free
Perfect Aircrete, Kitchen Ingredients.
Futuristic pixel-raising display lets you feel what's onscreen
Cutting-Edge Facility Generates Pure Water and Hydrogen Fuel from Seawater for Mere Pennies
This tiny dev board is packed with features for ambitious makers
Scientists Discover Gel to Regrow Tooth Enamel
Vitamin C and Dandelion Root Killing Cancer Cells -- as Former CDC Director Calls for COVID-19...
Galactic Brain: US firm plans space-based data centers, power grid to challenge China
A microbial cleanup for glyphosate just earned a patent. Here's why that matters
Japan Breaks Internet Speed Record with 5 Million Times Faster Data Transfer
AI AlphaGo Zero started from scratch to become best at Chess, Go and Japanese Chess within hours
Starting from random play, and given no domain knowledge except the game rules, AlphaZero achieved within 24 hours a superhuman level of play in the games of chess and shogi (Japanese chess) as well as Go, and convincingly defeated a world-champion program in each case.
Self-play games are generated by using the latest parameters for this neural network, omitting
the evaluation step and the selection of best player.
AlphaGo Zero tuned the hyper-parameter of its search by Bayesian optimization. In AlphaZero they reuse the same hyper-parameters for all games without game-specific tuning. The sole exception is the noise that is added to the prior policy to ensure exploration; this is scaled in proportion to the typical number of legal moves for that game type.
Like AlphaGo Zero, the board state is encoded by spatial planes based only on the basic
rules for each game. The actions are encoded by either spatial planes or a flat vector, again
based only on the basic rules for each game.