A cascade computer model for mocrobicide diffusivity from mucoadhesive formulations

摘要: The cascade computer model (CCM) was designed as a machine-learning feature platform for prediction of drug diffusivity from the mucoadhesive formulations. Three basic models (the statistical regression model, K nearest neighbor and modified version back propagation neural network) in CCM operate sequentially close collaboration with each other, employing estimated value obtained afore-positioned base an input to next-positioned cascade. effects various parameters on pharmacological efficacy female controlled delivery system (FcDDS) intended prevention women HIV-1 infection were evaluated using vitro apparatus “Simulant Vaginal System” (SVS). We used simulations explicitly examine changes FcDDS determine prognostic potency variable vivo formulation efficacy. results approach compared those individual multiple model. significantly lowered percentage mean error (PME) enhanced r2 values models. It noted that generated PME 21.82 at 48169 epoch iterations, which is improved 29.91 % 118344 epochs by network this study indicated sequential ensemble classifiers allowed accurate domain variance considerably reduces time required training phase. accurate, easy operate, cost-effective, thus, can serve valuable tool may yield new insights into understanding how drugs are diffused carrier systems exert their efficacies under clinical conditions.