d-lowl/kglsm

Hi there 👋

kglsm's Issues

Add documentation and examples

Example: SimAnneal for simple function

Cost function diffs

Allow for diffs in cost function: instead of calculating a cost function itself, calculate difference in cost between two solutions. Might decrease constant computation term, which is important for large instances of TSP specifically

Continuous Search Space sampling

Multiple sampling modes:

• Uniform, whole space
• Uniform, fixed neighbourhood
• Gaussians

recombination and mutation operators for permutation search spaces

To allow Evolutionary Algorithms on TSP

Multistep Delayed Annealing Schedule

Add option for simulated annealing routine to stay at the same temperature for multiple iterations.

Memory Efficient Serialisation of Memory

Add state machine composition

This is to add various binary operations on state machines/graphs to combine GLSMs

Add other prototypical search spaces

• Continuous
• 2-swap paths
• Constructive paths

Add multi-state SLS

• ILS with Restart
• Memetic Algorithm
• Random Iterative Improvement

Add those both as pre-built exploitable classes and as the code in the notebooks

Add CI

DSL

DSL for defining GLSMs outside the code. Potential fallback to JSON/YAML

Implement other model problems

TSP, Scheduling, Knapsack

Cannot create sequential LSM with single state

sequentialStateMachine<Double, BasicSolution<Double>> {
            step {
                name = "RandomWalk"
                strategy = RandomWalkStrategy(true)
            }
            terminationPredicate = FixedStepsPredicate(10)
        }

raises an exception

Exception in thread "main" java.lang.IllegalStateException: Required value was null.
	at space.d_lowl.kglsm.sls.StateMachineTransition$Builder.build(StateMachine.kt:36)
	at space.d_lowl.kglsm.sls.SequentialStateMachineBuilder.build(SequentialStateMachineBuilder.kt:109)

Add more sophisticated SLS strategies

TBC

• Population based, e.g. Genetic Algorithms
• Better ILS, e.g. Simulated Annealing

Check GLSM state machine for correctness

• Must terminate (eventually potentially reach termination state)
• Must be fully connected
• Must not have transitions from the final state

Logging and progress bars

Visualise GLSM state machine

Unify packages structure

TSP Model Problem

• Permutation search space
• 2-exchange neighbourhood
• TSP cost function
• Loading from file -- TSPlib

Optionals:

• k-exchange neighbourhood
• recombination and mutation operators for permutation search spaces

Simulated Annealing

Different temperature schedules

• Linear
• Multiplicative
• Slow
  - [ ] Multiple steps -- same temperature

External bindings

Python and binary apps. Potentially R?

Implement Iterated Local Search

Starting with SAT and binary neighbourhood?

Constructive search

So far only perturbative search has been considered

Refactor state machines

Setup kotlin notebooks with examples

Add memory event handlers to memory attributes

k-exchange neighbourhood

Investigate: solution as kmath vectors

Implement if feasible and appropriate

Simplify Generic Signatures

Current generic signatures are really cumbersome pretty much everywhere. We want as much of automatic inference as possible.

Add State Machine for strategies

Dokka

• Add KDoc/Dokka
• Host at jitpack

SLS Performance analysis

Emperical analysis, search space structure. Chapters 4, 5