Figure 1 | Scientific Reports

Figure 1

From: Automated machine learning for endemic active tuberculosis prediction from multiplex serological data

Figure 1

Schematic of the automated-machine learning platform MILO. Overview of the infrastructure and process for data processing, feature selection, and subsequent model training, building, initial validation, generalization testing and selection. MFI values for 31 anti-M.tb. antigens generated by multiplex microbead immunoassays comprise the balanced training dataset (Dataset A in this study). A large number of optimized models (> 300,000) were generated from the training dataset after data processing, feature selection, training, and validation. The true performance of the optimized models is then evaluated on the out-of-sample generalization (ideally prevalence-based) dataset (Datasets B and C in this study).

Back to article page