Application of data-intelligence algorithms for modeling the compaction performance of new pharmaceutical excipients

Abstract Background Pharmaceutical excipient development is an extensive process requiring a series of pre-formulation studies to evaluate their performance. The present study compares the conventional compaction and compression with artificial intelligence (AI) modeling predict performances thermally chemically modified starches obtained from Livingstone potato. Results native starch was by three methods, we following starches; pregelatinized (PS), ethanol dehydrated (ES), acid hydrolyzed (AS). Microcrystalline cellulose (Avicel® PH101) employed as reference since its use in tablet direct has been established. role pressure on degree volume reduction tablets studied using Kawakita Heckel models which highlighted that when pregelatinization followed dehydration, and/and or hydrolysis modification, directly compressible can be plastically deform. data-intelligence results indicated reliability AI-based over linear models. Hence, comparative demonstrated Adaptive neuro-fuzzy inference system (ANFIS) outperformed other two performance all four excipients considerable accuracy. Conclusion compressibility indices have shown matching characteristics AS ES Avicel® PH101 terms potential than PS. Moreover, data demonstrates satisfactory ANFIS hybrid model increased skills properties these novel pharmaceutical excipients.