> ML_ALGORITHM // SIMULATED-ANNEALING-SA_v1.0
Simulated Annealing (SA)
Probabilistic global optimization metaheuristic that mimics the thermodynamic cooling of heated metals to escape local minima.
Thermodynamic Metaheuristicsevolutionary-searchhigh-intrinsicsmall (<1k)
Back to All AlgorithmsComputational Complexity
Training Complexity:O(cooling_steps * iterations_per_temp)
Inference Complexity:O(1)
Hardware Profile
CPU Friendly:Yes
Requires GPU:No
Memory Footprint:low
Interpretability & Data
Interpretability Tier:high-intrinsic
Training Data Needs:small (<1k)
Interpretability Assessment
Single-agent trajectory with explicit temperature decay mirroring physical metallurgical annealing.
Suitable Tasks & Supported Modalities
Suitable Tasks:
combinatorial optimizationlayout planningchip floorplanning
Supported Modalities:
tabular
Implementing Libraries
SciPySciPy Community / NumFOCUS · v1.14.1
View Specsimanneal
Foundational Literature
Common Pitfalls & Warnings
- Cooling too quickly (quenching) freezes the system in poor local minima before adequate state space exploration