Chapter 12 - Integrated machine learning framework for computer-aided chemical product design

Abstract

Chemical industry is continuously looking for opportunities to manufacture the necessary commodity chemicals as well as to convert them into higher value-added chemicals-based products. Current global business environment encourages a short time-to-market for any potential chemical product. With the help of computer-aided product design methods, the R&D cycle of new products is able to be shortened, the financial resource as well as manpower can be saved. Toward current challenges existing in chemical product design methods, machine learning (ML) is regarded as a promising solution technique for computer-aided property prediction and product design. In this chapter, first, the integrated ML framework for computer-aided molecular design (CAMD) is presented. Then, the ML-CAMD framework is discussed in detail for the establishment of ML models for property prediction as well as chemical product design. Finally, two case studies are presented for the application of the proposed framework.