Development of Robust Data Computing Methodology (RDCM) for a Multidisciplinary Pharmaceutical Process Design

While data computing and analysis has seen significant advance in the type of analytical tools available, there are limitations in the pre-treatment of raw data. In many pharmaceutical industrial situations these often include a number of missing values. In addition, the quality characteristics of drug products are often multidisciplinary (i.e., not of the same type). In order to address these limitations, the main purpose of this paper is to propose a new robust data computing methodology (RDCM). RDCM can systemically estimate observed missing values by reducing the dimensionality of large pharmaceutical data sets. It can also incorporate multidisciplinary pharmaceutical situations. The primary objectives of this paper are threefold. First, we develop a robust data mining (RDM) procedure using an expectation maximization (EM) algorithm and a correlation-based feature selection (CBFS) method. Second, we propose a multidisciplinary optimization model for the optimal design of a pharmaceutical process by developing a multivariate robust design model using a nonlinear goal programming. Finally, our numerical example clearly shows that the proposed RDCM can efficiently be applied to a pharmaceutical process design.