Applications of Machine Learning in Bioprocess Development and Optimization

Abstract

Machine learning can play a role in the modeling and optimization of bioprocesses. Process modeling is required for parameter optimization, predicting relationships between critical parameters, searching decisive quality attributes, bioprocess implementation, and execution of experiments. Machine learning is already playing an important role in many bioprocesses such as drug discovery, vaccine development, protein therapeutics, etc. Machine-learning algorithms can accelerate bioprocess control, design, development, monitoring, and optimization. These bioprocesses depend on indefinite time-variant reaction kinetics, and the interaction of many variables with different degrees of correlation and time-oriented bioprocess responses. These complexities present a variety of novel challenges in bioprocess development, to which machine learning-based process modeling can represent a trustworthy solution. This chapter investigates various machine-learning tools and algorithms that can help improve bioprocess development and optimization.