Feature Ranking for SimBiology

Rank importance of parameters in a SimBiology model for parameter estimation.

Feature Ranking

Copyright 2023 The MathWorks, Inc.

Rank importance of parameters in a SimBiology model for parameter estimation. FeatureRanking computes

• MRMR (minimum redundancy maximum relevance, [1])
• NCA (neighborhood component analysis, [2])
• F-TEST based feature ranking.

The ranking scores are computed from parameter samples and associated model simulations. The parameter samples are generated using a SimBiology.Scenarios object, the associated model simulations are computed using a SimBiology.SimFunction.Getting Started

Read and run driver.mlx for more details for how to use compute and plot ranking scores.

Help

After installing this MATLAB Add-On, you can also type

>> help FeatureSelection

>> help FeatureSelection.histogram

in the MATLAB Command Window to get instructions for of how to use FeatureSelection.

Requirements

This feature is supported in MATLAB® versions R2022a and later. Required products are SimBiology® and Statistics And Machine Learning Toolbox™.

References

[1] Ding C. and Peng H., Minimum redundancy feature selection from microarray gene expression data. Journal of Bioinformatics and Computer Biology 2005; 3(2):185–205.

[2] Yang W., Wang K. and Zuo. W., Neighborhood component feature selection for high-dimensional data. Journal of Computers 2012; 7(1):161-168.