Feature Selection and Preprocessing

How it works

Preparation

Highlighted claims

• Dimensionality reduction will use Boruta and LASSO within each outer-loop training set. — Development and cross-site validation of machine-learning models for diagnosis and prognosis of stable angina with and without obstructive coronary artery disease: a study protocol
• The final modelling feature set will combine the features selected by Boruta and LASSO. — Development and cross-site validation of machine-learning models for diagnosis and prognosis of stable angina with and without obstructive coronary artery disease: a study protocol
• A directed acyclic graph will make assumed causal pathways, bias sources, dataset shift, measurement practices, and downstream treatment decisions explicit. — Development and cross-site validation of machine-learning models for diagnosis and prognosis of stable angina with and without obstructive coronary artery disease: a study protocol
• Training and testing subsets will be imputed separately to avoid data leakage. — Development and cross-site validation of machine-learning models for diagnosis and prognosis of stable angina with and without obstructive coronary artery disease: a study protocol
• Candidate features will be filtered by availability and expert domain input before dimensionality reduction. — Development and cross-site validation of machine-learning models for diagnosis and prognosis of stable angina with and without obstructive coronary artery disease: a study protocol
• Missing data will be imputed with MICE under a missing-at-random assumption. — Development and cross-site validation of machine-learning models for diagnosis and prognosis of stable angina with and without obstructive coronary artery disease: a study protocol