Reinforcement Learning Agent Design and Optimization with Bandwidth Allocation Model

• DOI: 10.36227/techrxiv.22118471.v1
• Publication Date: 2023-02-22T16:41:35Z
• Authors: Rafael F. Reale; Joberto Martins
• Keywords: Heuristics; Agent-Based Model

<p>Reinforcement learning (RL) is currently used in various real-life applications. RL-based solutions have the potential to generically address problems, including ones that are difficult solve with heuristics and meta-heuristics and, addition, set of problems issues where some intelligent or cognitive approach required. However, reinforcement agents require a not straightforward design important issues. RL agent include target problem modeling, state-space explosion, training process, efficiency. Research addresses these aiming foster dissemination. A BAM model, summary, allocates shares resources users. There three basic models several hybrids differ how they allocate share among This paper issue an The agent's objective investigates model can contribute AllocTC-Sharing (ATCS) analytically described simulated evaluate it mimics operation ATCS offload computational tasks from agent. essential argument researched whether algorithms integrated facilitate optimize its execution. analytical simulation presented demonstrate offloads assists optimization. </p>