Properties (60)
| Predicate | Object |
|---|---|
| gptkbp:instanceOf |
gptkb:physicist
|
| gptkbp:associated_with |
gptkb:Deep_Reinforcement_Learning
AI ethics discussions AI safety research future AI applications AI policy considerations |
| gptkbp:basedOn |
Monte Carlo Tree Search
|
| gptkbp:competesIn |
human players
other_AI |
| gptkbp:completed |
superhuman performance
|
| gptkbp:developedBy |
DeepMind
|
| gptkbp:diedIn |
Go
chess shogi |
| gptkbp:exhibits |
tournaments
|
| gptkbp:has |
strategic depth
training efficiency high computational power tactical strength no human knowledge input |
| https://www.w3.org/2000/01/rdf-schema#label |
AlphaZero
|
| gptkbp:influenced |
gptkb:AlphaGo_Zero
further AI research |
| gptkbp:is |
algorithm
self-learning general-purpose model-free zero knowledge |
| gptkbp:is_featured_in |
scientific publications
media articles |
| gptkbp:is_known_for |
high adaptability
strategic innovation dynamic learning efficient learning |
| gptkbp:is_part_of |
AI advancements
DeepMind's research portfolio AI game theory AI strategy development machine learning evolution |
| gptkbp:is_recognized_for |
breakthrough in AI
|
| gptkbp:is_studied_in |
live matches
|
| gptkbp:is_used_in |
research
AI competitions educational purposes game development performance evaluation AI conferences training_AI_models |
| gptkbp:learnsMove |
self-play
|
| gptkbp:led_to |
new algorithms
improvements in game AI |
| gptkbp:performance |
gptkb:AlphaGo
Elmo Stockfish |
| gptkbp:relatedTo |
gptkb:traditional_AI
human grandmasters |
| gptkbp:releasedIn |
2017
|
| gptkbp:training |
game data
|
| gptkbp:uses |
neural networks
reinforcement learning |