Greedy Optimization is an iterative problem-solving strategy in which locally optimal decisions are made at each step — such as selecting the single best beacon placement or parameter configuration at each iteration — without revisiting or reconsidering prior choices, with the aim of efficiently approximating a globally optimal solution. In WiFi/CSI and indoor sensing research, it matters because many deployment and calibration problems (e.g., beacon placement for crowd monitoring, parameter tuning for crowd simulation models) involve large, combinatorial search spaces where exhaustive search is computationally intractable, making greedy approaches a practical and often effective alternative. Key variants include incremental greedy construction, as used in adaptive beacon deployment frameworks like EABeD where beacons are added one at a time to maximize coverage gain, and greedy search within broader optimization pipelines where it serves as a component of parameter estimation routines applied to crowd simulation evaluation.
Dictionary term