ReliefF and SVM Example
This repository was created for anybody interested in using feature selection (ReliefF, Matlab: relieff) and support vector machines (SVM, Matlab: fitcsvm) as a minimum working example to reproduce steps described in the publication below (Doerr2020). Data is provided in the sub-folder '_Data'. Structural features were extracted from micro-X-ray tomography data. ReliefF and SVM were used to build a classifier for the detection of broken pharmaceutical pellets within the sample.
Input Data:
(1) Extracted features of six ibuprofen (IBU) capsules (1763 pellets, 206 features):
'Desc_DataFile_C0.csv'
'Desc_DataFile_C1.csv'
'Desc_DataFile_C2.csv'
'Desc_DataFile_C3.csv'
'Desc_DataFile_C4.csv'
'Desc_DataFile_C5.csv'
(2) User defined feature categories:
'Feature_Categories.csv'
(3) Results of a feature sensitivity analysis:
'Feature_SenAnlys_Score.csv'
%------------------------------------------------------------------------------------------------
% Code written by Frederik Doerr, Feb 2020 (MATLAB R2019b)
% Application: For 'Support Vector Machine - Introduction and Application'
% % % Reference (open access):
% Doerr, F. J. S., Florence, A. J. (2020)
% A micro-XRT image analysis and machine learning methodology for the characterisation of multi-particulate capsule formulations.
% International Journal of Pharmaceutics: X.
% https://doi.org/10.1016/j.ijpx.2020.100041
% Data repository: https://doi.org/10.15129/e5d22969-77d4-46a8-83b8-818b50d8ff45
% Video Abstract: https://strathprints.strath.ac.uk/id/eprint/71463
%------------------------------------------------------------------------------------------------
Cita come
Doerr, Frederik J. S., and Alastair J. Florence. “A Micro-XRT Image Analysis and Machine Learning Methodology for the Characterisation of Multi-Particulate Capsule Formulations.” International Journal of Pharmaceutics: X, vol. 2, Elsevier BV, Dec. 2020, p. 100041, doi:10.1016/j.ijpx.2020.100041.
Compatibilità della release di MATLAB
Compatibilità della piattaforma
Windows macOS LinuxTag
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!Scopri Live Editor
Crea script con codice, output e testo formattato in un unico documento eseguibile.
Le versioni che utilizzano il ramo predefinito di GitHub non possono essere scaricate
Versione | Pubblicato | Note della release | |
---|---|---|---|
1.0.1 | Minor corrections in description, references |
|
|
1.0.0 |
|