An open API service indexing awesome lists of open source software.

https://github.com/ncordon/feature-selection

Repo de resolución de prácticas de la asignatura Metaheurísticas, UGR
https://github.com/ncordon/feature-selection

Last synced: 4 months ago
JSON representation

Repo de resolución de prácticas de la asignatura Metaheurísticas, UGR

Awesome Lists containing this project

README

          

# Feature Selection

Contains implementations from scratch of several metaheuristics adjusted for feature selection, and its application to three
different datasets:

- [Wdbc](data/wdbc.arff)
- [Arrhythmia](data/arrhythmia.arff)
- [Movement Libras](data/movement_libras.arff)

## License
GNU licensed

## Files

* `bin` contains the implemented code.
* Simulated Annealing `ES.r`
* Local Search `BL.r`
* Tabu Search: `BT.r`, `BText.r`
* Basic Multirestart `BMB.r`
* Genetic algorithms `AGs.r`
* GRASP `GRASP.r`
* ILS `ILS.r`
* Ant Colony Optimization `OCH.r`
* Memetic algorithms `AGs.r`
* `data` contains datasets, graphs and boxplots.
* `pscode` contains pseudocodes that are compiled to `.tex` doing `make`
* `tex` contains documentation, only available in Spanish:
* `trajectories.tex`
* `multi-restart.tex`
* `genetic.tex`
* `och.tex`
* `memetic.tex`

## Usage

Execute `src/main.r` to get the results
of each algorithm.

For example, to gather the results from local search algorithm applied to all three datasets:
```r
BL.results <- cross.eval(BL)
```

`params.r` contains settings used by the algorithms.

![Boxplot del dataset Wdbc](data/trajectories/wdbc_3nn_boxplot.png)