Ask a Question

Prefer a chat interface with context about you and your work?

Beyond algorithm hyperparameters: on preprocessing hyperparameters and associated pitfalls in machine learning applications

Beyond algorithm hyperparameters: on preprocessing hyperparameters and associated pitfalls in machine learning applications

Adequately generating and evaluating prediction models based on supervised machine learning (ML) is often challenging, especially for less experienced users in applied research areas. Special attention is required in settings where the model generation process involves hyperparameter tuning, i.e. data-driven optimization of different types of hyperparameters to improve the predictive …