Package for Multi-Criteria Decision Analysis with Preference Identification
Project description
EVO-SPOTIS
This is Python 3 library for multi-criteria decision analysis with decision-maker preference identification.
Installation
Downloading and installation of evo_spotis
package can be done using pip
pip install evo-spotis
Methods
mcda_methods
module includes:
spotis
with SPOTIS method (the Stable Preference Ordering Towards Ideal Solution method)
stochastic_algorithms includes:
- DE algorithm
DE_algorithm
(the Differential Evolution algorithm)
The DE algorithm is applied for the identification of criteria weights (decision-maker preferences) based on a training dataset with evaluated alternatives,
including alternatives performances (training features) and their ranking (target variable). The goal (fitness) function uses the correlation coefficient
of predicted ranking with real ranking. The predicted ranking is generated using the SPOTIS method and weights calculated by the DE algorithm in each DE iteration.
It is a profit function. Therefore, higher values denote better results. Examples of use of evo_spotis
are included on GitHub in examples
Other modules:
additions
includingrank_preference
method for ranking alternatives according to MCDA scorecorrelations
containing:- Spearman rank correlation coefficient
spearman_coeff
, - Weighted Spearman rank correlation coefficient
weighted_spearman_coeff
, - Pearson correlation coefficient
pearson_coeff
- Spearman rank correlation coefficient
normalizations
with methods for decision matrix normalization:linear_normalization
- Linear normalization,minmax_normalization
- Minimum- Maximum normalization,max_normalization
- Maximum normalization,sum_normalization
- Sum normalization,vector_normalization
- Vector normalization
weighting_methods
containing:entropy_weighting
- Entropy objective weighting method.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for evo_spotis-0.0.5-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 12c52893a8a7d1a26855001388d5abced09af523721458a90105c80c08ed1b8b |
|
MD5 | e6346ada1d8481af58b0ffc08f08bfce |
|
BLAKE2b-256 | a198f879f1da660b7e9c046c31be1db65436be605fba263b06932dfe9bef1b4c |