7 results on '"Macromanagement"'
Search Results
2. Learning Macromanagement in Starcraft by Deep Reinforcement Learning
- Author
-
Wenzhen Huang, Qiyue Yin, Junge Zhang, and Kaiqi Huang
- Subjects
StarCraft ,macromanagement ,Asynchronous Advantage Actor-Critic ,Chemical technology ,TP1-1185 - Abstract
StarCraft is a real-time strategy game that provides a complex environment for AI research. Macromanagement, i.e., selecting appropriate units to build depending on the current state, is one of the most important problems in this game. To reduce the requirements for expert knowledge and enhance the coordination of the systematic bot, we select reinforcement learning (RL) to tackle the problem of macromanagement. We propose a novel deep RL method, Mean Asynchronous Advantage Actor-Critic (MA3C), which computes the approximate expected policy gradient instead of the gradient of sampled action to reduce the variance of the gradient, and encode the history queue with recurrent neural network to tackle the problem of imperfect information. The experimental results show that MA3C achieves a very high rate of winning, approximately 90%, against the weaker opponents and it improves the win rate about 30% against the stronger opponents. We also propose a novel method to visualize and interpret the policy learned by MA3C. Combined with the visualized results and the snapshots of games, we find that the learned macromanagement not only adapts to the game rules and the policy of the opponent bot, but also cooperates well with the other modules of MA3C-Bot.
- Published
- 2021
- Full Text
- View/download PDF
3. Comparative Analysis of Micro - and Macromanagement Features of the Inclusive School.
- Author
-
Bucăloiu, Ionela
- Subjects
CLASSROOM management ,COMPARATIVE studies ,SCHOOL administration ,SCHOOLS ,TEACHERS - Abstract
Making an effective school and classroom management in particular is indeed a challenge for both the director and the teachers. Because, it is necessary to consider the action of continuous regulation and efficiency of the communicative process, transformation of the institution into an inclusive organization in the sense of knowing, understanding the differences that arise from the multiple interactions between the people involved in the educational process, irrespective of the hierarchical line, but especially in order to exploit and maximize these differences in the direction of increasing the efficiency and productivity of work at both individual and collective level. [ABSTRACT FROM AUTHOR]
- Published
- 2019
4. Land Institutions and Chinese Political Economy.
- Author
-
Rithmire, Meg Elizabeth
- Subjects
- *
LAND use , *ECONOMIC development , *LAND management , *ECONOMIC reform , *HISTORY ,ECONOMIC conditions in China - Abstract
This article critically examines the origins and evolution of China’s unique land institutions and situates land policy in the larger context of China’s reforms and pursuit of economic growth. It argues that the Chinese Communist Party (CCP) has strengthened the institutions that permit land expropriation—namely, urban/rural dualism, decentralized land ownership, and hierarchical land management—in order to use land as a key instrument of macroeconomic regulation, helping the CCP respond to domestic and international economic trends and manage expansion and contraction. Key episodes of macroeconomic policymaking are analyzed, with the use of local and central documents, to show how the CCP relied on the manipulation and distribution of the national land supply either to stimulate economic growth or to rein in an overheating economy. China’s land institutions, therefore, share “complementarities” with fiscal and financial institutions and benefit powerful political actors while imposing costs on marginal ones. [ABSTRACT FROM AUTHOR]
- Published
- 2017
- Full Text
- View/download PDF
5. Learning Macromanagement in Starcraft by Deep Reinforcement Learning
- Author
-
Kaiqi Huang, Qiyue Yin, Junge Zhang, and Wenzhen Huang
- Subjects
0209 industrial biotechnology ,Computer science ,02 engineering and technology ,TP1-1185 ,ENCODE ,Biochemistry ,Article ,Analytical Chemistry ,020901 industrial engineering & automation ,0202 electrical engineering, electronic engineering, information engineering ,Reinforcement learning ,Electrical and Electronic Engineering ,Instrumentation ,Queue ,business.industry ,Chemical technology ,Perfect information ,ComputingMilieux_PERSONALCOMPUTING ,Variance (accounting) ,macromanagement ,Atomic and Molecular Physics, and Optics ,Asynchronous Advantage Actor-Critic ,Recurrent neural network ,StarCraft ,Asynchronous communication ,020201 artificial intelligence & image processing ,Macromanagement ,Artificial intelligence ,business - Abstract
StarCraft is a real-time strategy game that provides a complex environment for AI research. Macromanagement, i.e., selecting appropriate units to build depending on the current state, is one of the most important problems in this game. To reduce the requirements for expert knowledge and enhance the coordination of the systematic bot, we select reinforcement learning (RL) to tackle the problem of macromanagement. We propose a novel deep RL method, Mean Asynchronous Advantage Actor-Critic (MA3C), which computes the approximate expected policy gradient instead of the gradient of sampled action to reduce the variance of the gradient, and encode the history queue with recurrent neural network to tackle the problem of imperfect information. The experimental results show that MA3C achieves a very high rate of winning, approximately 90%, against the weaker opponents and it improves the win rate about 30% against the stronger opponents. We also propose a novel method to visualize and interpret the policy learned by MA3C. Combined with the visualized results and the snapshots of games, we find that the learned macromanagement not only adapts to the game rules and the policy of the opponent bot, but also cooperates well with the other modules of MA3C-Bot.
- Published
- 2021
- Full Text
- View/download PDF
6. Learning Macromanagement in Starcraft by Deep Reinforcement Learning.
- Author
-
Huang, Wenzhen, Yin, Qiyue, Zhang, Junge, Huang, Kaiqi, Šumak, Boštjan, and Pušnik, Maja
- Subjects
- *
REINFORCEMENT learning , *STRATEGY games , *DEEP learning , *RECURRENT neural networks , *RULES of games , *ARTIFICIAL intelligence - Abstract
StarCraft is a real-time strategy game that provides a complex environment for AI research. Macromanagement, i.e., selecting appropriate units to build depending on the current state, is one of the most important problems in this game. To reduce the requirements for expert knowledge and enhance the coordination of the systematic bot, we select reinforcement learning (RL) to tackle the problem of macromanagement. We propose a novel deep RL method, Mean Asynchronous Advantage Actor-Critic (MA3C), which computes the approximate expected policy gradient instead of the gradient of sampled action to reduce the variance of the gradient, and encode the history queue with recurrent neural network to tackle the problem of imperfect information. The experimental results show that MA3C achieves a very high rate of winning, approximately 90%, against the weaker opponents and it improves the win rate about 30% against the stronger opponents. We also propose a novel method to visualize and interpret the policy learned by MA3C. Combined with the visualized results and the snapshots of games, we find that the learned macromanagement not only adapts to the game rules and the policy of the opponent bot, but also cooperates well with the other modules of MA3C-Bot. [ABSTRACT FROM AUTHOR]
- Published
- 2021
- Full Text
- View/download PDF
7. Macromanagement of ITS and Logistics Services Demand and Supply
- Author
-
Radačić, Željko, Bošnjak, Ivan, Čurepić, Danko, and Fabjan, Daša
- Subjects
Logistics ,Intelligent Transport Systems ,Macromanagement ,Interactions - Abstract
The development and deployment of Intelligent Transport Systems and logistics services will ultimately need to be integrated within larger transport and communications policies and programs on the state level. Although the basic tehnical standards and the European ITS architectural framework are ensured, several macromanagement issues are open. In our environment transport telematics and ITS were initially deployed as a set of semi-autonomous local implementations with different actors from the transport, communications and computing industries. The development of ITS and logistics services are evolutionary and their implementation relies on their interactions with other services and potentionally with services outside the ITS domain. The national ITS architecture and Intelligent Transport City models (such is CROCITS) define technological and beyond-technological frameworks within which the ITS user service can be successfully implemented. Effective planning and coordination of logistics and implementation have to ensure: - prevention of undesirable side effects, - system coherency and interoperability, - public/private partnership. The paper will consider the ITS macro-management as a set of measures and instruments aimed at effective demand and supply management of ITS and logistics services. Within the ITS architectural framework and the accepted ITS program, specialized ITS activities can be harmonized among actors. The paper will also analyze the direct and indirect "system interactions" between the end-users and the operators.
- Published
- 2003
Catalog
Discovery Service for Jio Institute Digital Library
For full access to our library's resources, please sign in.