• Nishant Agarwal
• Abhishek Kumar
• Achintya Singhal

This paper presents an overview of the development of bots for competitive real-time strategy (RTS) games using Artificial Intelligence. A comparative study is made between the design of AI bots for strategic board games and RTS games and the challenges involved in the development of an AI bot for RTS games are highlighted. Existing work addressing these challenges are discussed. Important concepts used in designing an RTS game AI such as Imitation Learning, Adaptive AI and Reinforcement Learning (RL) are explained. An implementation of AI bot for the RTS game DOTA 2 (Defense of the Ancients 2) using Reinforcement Learning is presented and the “reward function” used in the implementation is discussed. Finally, the recent developments in board game AI (with respect to Chess and Go) and in RTS game AI (with respect to DOTA 2) are presented and the future prospects are discussed.

Artificial intelligence (AI), computer games, real-time strategy (RTS), multiplayer online battle arena (MOBA), defense of the ancients (DOTA), imitation learning, adaptive AI, reinforcement learning (RL), task allocation (TA)

"Artificial Intelligence in Real-Time Strategy Games", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.6, Issue 4, page no.443-450, April-2019, Available :http://www.jetir.org/papers/JETIR1904266.pdf

"Artificial Intelligence in Real-Time Strategy Games", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.6, Issue 4, page no. pp443-450, April-2019, Available at : http://www.jetir.org/papers/JETIR1904266.pdf