This repository contains the implementation and evaluation of traditional (GLM, Lasso and Elastic Net) and non-traditional (Tree Boosting, SVM, Naive Bayes and Multilayer Perceptron) machine learning algorithms for classification. Additionally, for each computed model explanations (Shapley values for all, weights for GLM, Taylor Decomposition for MLP) are calculated and visualized. Classification performance is evaluated using 10 different measures.
The framework has been applied in these publications:
Dengler, Nora Franziska, Vince Istvan Madai, Meike Unteroberdörster, Esra Zihni, Sophie Charlotte Brune, Adam Hilbert, Michelle Livne, Stefan Wolf, Peter Vajkoczy, and Dietmar Frey. 2021. “Outcome Prediction in Aneurysmal Subarachnoid Hemorrhage: A Comparison of Machine Learning Methods and Established Clinico-Radiological Scores.” Neurosurgical Review https://link.springer.com/article/10.1007/s10143-020-01453-6 (open access)
Zihni, Esra, Vince Istvan Madai, Michelle Livne, Ivana Galinovic, Ahmed A. Khalil, Jochen B. Fiebach, and Dietmar Frey. 2020. “Opening the Black Box of Artificial Intelligence for Clinical Decision Support: A Study Predicting Stroke Outcome.” PLOS ONE https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0231166 (open access)
Manual to this framework can be found here.
This project is licensed under the MIT license.
Esra Zihni, Adam Hilbert