Application of a kNN‐based similarity method to biopharmaceutical manufacturing

Abstract

Machine learning‐based similarity analysis is commonly found in many artificial intelligence applications like the one utilized in e‐commerce and digital marketing. In this study, a kNN‐based (k‐nearest neighbors) similarity method is proposed for rapid biopharmaceutical process diagnosis and process performance monitoring. Our proposed application measures the spatial distance between batches, identifies the most similar historical batches, and ranks them in order of similarity. The proposed method considers the similarity in both multivariate and univariate feature spaces and measures batch deviations to a benchmarking batch. The feasibility and effectiveness of the proposed method are tested on a drug manufacturing process at Biogen.

Technology outlook for real‐time quality attribute and process parameter monitoring in biopharmaceutical development—A review

“…K‐means clustering (grouping observations into categories based on their distance to centroids) and hierarchical clustering (presenting distances between observations in a tree‐like structure) are among the simplest and most often used modeling approaches. Ren et al (2019) have demonstrated the use of the K‐means clustering method to identify the similarities of historical pharmaceutical batches. Data‐driven advanced analytics for process modeling (Kornecki & Strube, 2019), process design space analysis (Bhatia et al, 2016), and the connection between CPPs and CQAs have been covered elsewhere (Sokolov et al, 2016; Sokolov et al, 2017).…”

Process analytics 4.0: A paradigm shift in rapid analytics for biologics development

“…Ren and coworkers reported the use of the K-means clustering method to identify the similarities of historical pharmaceutical batches. 85 This machine learning-based similarity method was proposed for rapid biopharmaceutical process diagnosis and process performance monitoring. 86 Data-driven process modeling and design space analysis are also pivotal applications of advanced data analytics.…”

“…Data‐driven, mechanistic, and/or hybrid modeling can be used based on the output requirements 6,10,84 . For example Ren and coworkers reported the use of the K‐means clustering method to identify the similarities of historical pharmaceutical batches 85 . This machine learning‐based similarity method was proposed for rapid biopharmaceutical process diagnosis and process performance monitoring 86 .…”

A Context‐Aware Assisted WiFi Positioning Method

“…Common core algorithms for location fingerprint location include k-nearest neighbor (KNN) algorithm, Bayesian Information Criterion (BIC), Support Vector Machine (SVM) algorithm, decision tree, and K-means. Among them, the KNN algorithm has been widely applied to such aspects as location [4], transportation [5], medical treatment [6,7], network [8][9][10], energy [11], and operations research [12]. At the same time, it is also quite advantageous to performance.…”