Authors | Date |
---|---|
Majid Afshar and Hamid Usefi |
01/10/20 |
The code of SVFS is available now, and a link to the paper is given. If you need more details and explanation about the algorithm, please contact Majid Afshar or Hamid Usefi.
Here is a link to the paper : https://www.nature.com/articles/s41598-021-83150-y
To determine the most important features using the algorithm described in "Dimensionality Reduction Using Singular Vectors" by Majid Afshar and Hamid Usefi
This code can be run using Python 3.2 and above. Also, the following packages should be installed in your environment as the program dependencies:
- Pandas
- Numpy
- Scikit learn
- Networkx
To run the code, open main.py
and specify a list of datasets to apply the method. We note the dataset does not have any headers (neither the features nor the samples IDs). You can add any high dimensional dataset to Datasets and insert their name in the list of datasets in 'main.py'.
All datasets must be stored in Datasets folder. As part of our experiments, we use datasets from Gene Expression Omnibus (GEO), and datasets can be cleaned by this code.