Muxstep: an open-source C ++ multiplex HMM library for making inferences on multiple data types.

Area of Science:

• Computational biology
• Machine learning
• Network science

Background:

• Advancements in experimental methods yield large, multi-dimensional datasets.
• Conventional machine learning struggles with high-dimensional and multi-type data.
• Complex networks with topological features offer a promising modeling approach.

Purpose of the Study:

• Introduce MUXSTEP, an open-source library for binary classification.
• Utilize multiplex networks to handle multi-type data effectively.
• Provide a flexible framework for advanced machine learning modeling.

Main Methods:

• Development of the MUXSTEP library.
• Application of multiplex network frameworks for data modeling.
• Binary classification tasks on diverse data types.

Main Results:

• Demonstration of MUXSTEP's utility for multi-type data classification.
• Establishment of a foundation for multiplex network-based machine learning.
• Creation of an adaptable library for various modeling needs.

Conclusions:

• MUXSTEP offers a novel approach to machine learning with complex, multi-dimensional data.
• The library facilitates the development of more expressive and accurate models.
• MUXSTEP is designed for both immediate use and custom modification.