exprso: an R-package for the rapid implementation of machine learning algorithms

Area of Science:

• Computational biology
• Bioinformatics
• Data science

Background:

• Machine learning is crucial in scientific research.
• Non-expert programmers face challenges implementing complex machine learning pipelines.
• Accurate and generalizable models are essential for scientific discovery.

Purpose of the Study:

• Introduce exprso, an R package designed for non-expert programmers.
• Provide an intuitive machine learning suite for high-dimensional data classification.
• Simplify the implementation of elaborate machine learning workflows.

Main Methods:

• Developed an object-oriented framework in R.
• Encapsulated common analytical methods into interchangeable modules.
• Included modules for feature selection, classification, parameter tuning, cross-validation, and prediction.

Main Results:

• exprso facilitates the classification of high-dimensional data.
• The package supports advanced techniques like Monte Carlo and nested cross-validation.
• It enables multi-class classification and continuous outcome prediction.

Conclusions:

• exprso lowers the barrier for non-expert programmers to utilize advanced machine learning.
• The package enhances the development of accurate and generalizable models in scientific investigations.
• exprso offers a flexible and comprehensive solution for diverse machine learning tasks.