Skip to main content

Universidade Federal do Rio Grande do Sul

Organization2 reports

2 reports, presentations and talks by Universidade Federal do Rio Grande do Sul in the Game Industry Library, the newest “A Taxonomy of Collectible Card Games: A Game-Playing AI Perspective” (2024).

Documents

Cover of A Taxonomy of Collectible Card Games: A Game-Playing AI Perspective
Whitepaper16 pages

A Taxonomy of Collectible Card Games: A Game-Playing AI Perspective

Collectible Card Games represent a complex domain for artificial intelligence research, characterized by a dual-phase structure consisting of strategic deck-building and tactical battling.

  • Collectible Card Games (CCGs) are defined by a dual-phase structure that separates strategic deck-building from tactical, in-game battling.
  • A formal taxonomy of ten representative CCGs reveals that core mechanical foundations—specifically resource management, turn structures, and win conditions—can be standardized for AI analysis.
  • Computational complexity in CCGs is directly proportional to the density of mechanical elements, such as card abilities and drafting formats, which dictate the difficulty of decision-making.
AIGame DesignCollectible Card Game+1
Universidade Federal de Minas GeraisSep 2024
Cover of Enhancing Deep Reinforcement Learning for Scale Flexibility in Real-Time Strategy Games
Whitepaper43 pages

Enhancing Deep Reinforcement Learning for Scale Flexibility in Real-Time Strategy Games

Real-time strategy games present a significant challenge for artificial intelligence due to the need for agents to navigate varying map dimensions and dynamic unit counts.

  • Integrating Spatial Pyramid Pooling (SPP) into reinforcement learning architectures enables AI agents to adapt to varying map sizes and unit counts without requiring neural network structural modifications.
  • This scale-invariant framework achieves superior performance compared to traditional baselines while reducing the required training budget by approximately 66%.
  • A training regimen utilizing sequential, high-frequency environment-swapping is essential for ensuring that agents develop robust, generalizable policies.
AIGame DevelopmentReal-Time Strategy+2
Universidade Federal de Minas GeraisJul 2024

Do you work at Universidade Federal do Rio Grande do Sul? Get a free report of how these documents are read and cited here.