Optimization by Simulated Annealing.¶
Kirkpatrick, S., Gelatt, C. D., & Vecchi, M. P. (1983). Optimization by Simulated Annealing. Science, 220(4598), 671-680.
Cited by¶
9 citations across 9 artifacts.
Each citation links to the sentence it supports in the citing article.
Primes¶
- Annealing
- Combinatorial optimisation: simulated annealing accepts uphill moves freely at high "temperature," then cools its acceptance threshold so the search settles near a global optimum instead of the nearest local one — the Metropolis criterion is a direct transcription of the metallurgical protocol.
This sourceTransfers the metallurgical heat-then-cool protocol to combinatorial optimization, with the Metropolis acceptance rule as the direct transcription of the statistical-mechanical process.
Supported in partVerified against the publisher's abstract
The abstract states the statistical-mechanics/annealing analogy behind simulated annealing but not the Metropolis acceptance rule, the cooling schedule, or global-optimum convergence; full text was not retrievable.
“There is a deep and useful connection between statistical mechanics (the behavior of systems with many degrees of freedom in thermal equilibrium at a finite temperature) and multivariate or combinatorial optimization (finding the minimum of a given function depending on many parameters). A detailed analogy with annealing in solids provides a framework for optimization of the properties of very large and complex …”
- Combinatorial optimisation: simulated annealing accepts uphill moves freely at high "temperature," then cools its acceptance threshold so the search settles near a global optimum instead of the nearest local one — the Metropolis criterion is a direct transcription of the metallurgical protocol.
- Eventual Realisation of Possibility
- Information and search: any codeword in a finite alphabet used long enough is eventually sent; branch-and-bound, simulated annealing, and evolutionary algorithms reach every connected region given enough exploration.
This sourceIntroduces simulated annealing as stochastic search; the method converges to a global optimum with probability one given sufficient exploration — any connected region is eventually reached.
Supported in partVerified against the publisher's abstract
The abstract frames simulated annealing as an optimization framework for very large, complex systems, but says nothing of reaching every connected region or of convergence guarantees.
“A detailed analogy with annealing in solids provides a framework for optimization of the properties of very large and complex systems.”
- Information and search: any codeword in a finite alphabet used long enough is eventually sent; branch-and-bound, simulated annealing, and evolutionary algorithms reach every connected region given enough exploration.
- Local Optimum
- The same transfer runs from engineering to personal decision-making, where simulated annealing's cooling schedule — high temperature early, low temperature late — ports as career advice (explore broadly early, exploit later) with the identical structural justification: early non-local search covers more basins before the cost of moving rises.
This sourceThe cooling schedule (high temperature early, low late) that escapes local optima by temporarily accepting worse states.
Supported in partVerified against the publisher's abstract
Gives the origin of simulated annealing and its annealing-as-optimization framework, but not the high-early/low-late cooling schedule and nothing about a port to personal decision-making or career advice.
“A detailed analogy with annealing in solids provides a framework for optimization of the properties of very large and complex systems.”
- The same transfer runs from engineering to personal decision-making, where simulated annealing's cooling schedule — high temperature early, low temperature late — ports as career advice (explore broadly early, exploit later) with the identical structural justification: early non-local search covers more basins before the cost of moving rises.
- Metastability
- The optimization community's intuition that a learner is "trapped in a local minimum," together with its engineering responses — simulated annealing, momentum, restart schedules — is the materials-science barrier-crossing picture ported into search.
This sourcePorts the materials-science barrier-crossing / annealing picture into combinatorial search to escape local minima.
Supported in partVerified against the publisher's abstract
The abstract states the statistical-mechanics/annealing-in-solids analogy as a framework for combinatorial optimization, backing the simulated-annealing half of the claim but not momentum or restart schedules.
“There is a deep and useful connection between statistical mechanics (the behavior of systems with many degrees of freedom in thermal equilibrium at a finite temperature) and multivariate or combinatorial optimization (finding the minimum of a given function depending on many parameters). A detailed analogy with annealing in solids provides a framework for optimization of the properties of very large and complex …”
- The optimization community's intuition that a learner is "trapped in a local minimum," together with its engineering responses — simulated annealing, momentum, restart schedules — is the materials-science barrier-crossing picture ported into search.
- Optimization Landscape
- Simulated annealing — the statistical-physics insight that slow cooling escapes local minima — transferred into combinatorial optimisation and then into stochastic-gradient methods in deep learning, the structural intuition being the same: increase exploration when topology is rugged.
This sourceImports the statistical-physics slow-cooling insight into combinatorial optimization to escape local minima.
Supported in partVerified against the publisher's abstract
The 1983 abstract backs only the statistical-mechanics-to-combinatorial-optimisation import; it says nothing about transfer to stochastic-gradient deep learning or a rugged-landscape exploration rule.
“There is a deep and useful connection between statistical mechanics (the behavior of systems with many degrees of freedom in thermal equilibrium at a finite temperature) and multivariate or combinatorial optimization (finding the minimum of a given function depending on many parameters). A detailed analogy with annealing in solids provides a framework for optimization of the properties of very large and complex …”
- Simulated annealing — the statistical-physics insight that slow cooling escapes local minima — transferred into combinatorial optimisation and then into stochastic-gradient methods in deep learning, the structural intuition being the same: increase exploration when topology is rugged.
- Simulated Annealing
- Kirkpatrick, Gelatt, and Vecchi (1983)
This sourceFoundational paper introducing simulated annealing as a general-purpose optimization metaheuristic; recasts the Metropolis algorithm with temperature-controlled probabilistic acceptance, demonstrated on TSP and VLSI placement.
Supported in partVerified against the publisher's abstract
The 1983 Science paper's own abstract confirms its statistical-mechanics and annealing-analogy framing of optimization, but states nothing about introducing simulated annealing or about TSP/VLSI demonstrations.
“There is a deep and useful connection between statistical mechanics (the behavior of systems with many degrees of freedom in thermal equilibrium at a finite temperature) and multivariate or combinatorial optimization (finding the minimum of a given function depending on many parameters). A detailed analogy with annealing in solids provides a framework for optimization of the properties of very large and complex …”
- Kirkpatrick, Gelatt, and Vecchi (1983)
Mechanisms¶
- Annealing, Noise, or Random Restart
- When the landscape is unknown and the current state is merely the best nearby, exploration is the only way forward.
This sourceScience. 220(4598): 671–680. (1983). Introduces simulated annealing as a way to explore complex cost landscapes and escape local optima that defeat purely local improvement.
- When the landscape is unknown and the current state is merely the best nearby, exploration is the only way forward.
- Annealing or Perturbation Schedule
- Its honest failure mode is the cooling schedule itself: cool too fast and you quench into a mediocre local optimum; cool too slow and you burn budget wandering long after the good basin was in reach.
This sourceExplains that annealing quality depends on the cooling schedule and that rapid quenching can trap a search in a nonoptimal state.
- Its honest failure mode is the cooling schedule itself: cool too fast and you quench into a mediocre local optimum; cool too slow and you burn budget wandering long after the good basin was in reach.
- Annealing Temperature Schedule
- The team's real work was choosing that curve — hot enough at the start to escape traps, slow enough to cool that good structure survives — because on a badly-set schedule the same solver either freezes into the first mediocre loop or never settles at all.
This sourceDefines simulated annealing as initially accepting disruptive moves at high temperature and then cooling gradually so exploration can escape local traps before the search settles; overly rapid or inadequate cooling defeats those aims.
- The team's real work was choosing that curve — hot enough at the start to escape traps, slow enough to cool that good structure survives — because on a badly-set schedule the same solver either freezes into the first mediocre loop or never settles at all.
Verification¶
Does it exist? Confirmed. This work's DOI resolves to a registered record, which fixes its identity. That is all it fixes.
Does it back the claim? Read against the text for 6 of 9 citations: 6 supported in part. Each verdict is shown under its citation below, with what in the work backs the sentence.
Was it audited? Yes. A second, independent pass read the citation against the article text and recorded a verdict.
Support is checked per citation rather than per work — the same source can be cited soundly in one article and wrongly in another. Per-citation recording began recently, so a citation with no recorded check is a gap in the record rather than evidence it went unchecked.
See how references were verified.
Registry ID ref:ecc97c33cc44 · see in the full table