Skip to content
/ GASOM Public

GASOM: Genetic Algorithm assisted Architecture Learning in Self Organizing Maps

License

Notifications You must be signed in to change notification settings

iMachLab/GASOM

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

6 Commits
 
 
 
 
 
 

Repository files navigation

GASOM

GASOM: Genetic Algorithm assisted Architecture Learning in Self Organizing Maps

Article detail:

The article has been accepted (to appear, or published) in the Proceedings of the 24th International Conference on Neural Information Processing, held in Guangzhou, China, November 14–18, 2017. (http://www.iconip2017.org/)

Authors:

Ashutosh Saboo, Anant Sharma, Tirtharaj Dash

Data Science Research Group, Department of Computer Science, BITS Pilani, Goa Campus, India

Citation request:

Will be updated soon.

Language:

The GASOM software is written in Python. The GA implementation of the software is parallelized for efficiency.

Dependencies:

  1. Pyevolve==0.6rc1
  2. matplotlib==2.0.0
  3. numpy==1.12.1
  4. pandas==0.18.1

Setup:

  1. pip install -r requirements.txt
  2. Edit the params : datasetpath, number_of_columns_csv, features, dataset_name, type_of_problem, data (Change data numpy array, so that, data contains only the relevant features, without the tags and indices)
  3. python train.py > dataset.log (This gives the best possible SOM Map Size for your dataset)
  4. Results will be present in dataset_name folder in cwd, along with final stats in dataset.log file.
  5. python generate_error_plot.py <dataset_name> (Error plot is generated)
  6. Visualise the results

Data sets used for testing our Model:

  • Real World Data sets used:

    1. Wine
    2. Iris
    3. Abalone
    4. Car Evaluation
    5. Glass Identification
    6. Sonar
  • Synthetic Data sets used:

    1. Corner
    2. CrescentFullMoon
    3. Ginger Breadman
    4. Half Kernal
    5. Outliers
    6. Two Spirals

About

GASOM: Genetic Algorithm assisted Architecture Learning in Self Organizing Maps

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published