[Retracted] A Novel Artificial Intelligence System in Formulation Dissolution Prediction

Table 1

The comparison of related projects.

Projects	ANN model	Regression method	Input and output	Amounts of parameters and samples
This project	ANN with backpropagation network	EDRM and RLRM	Input: the factors of tablets and the time points of dissolution. Output: dissolution results in every time points.	21 samples and 7 parameters. 3 kinds of formulation composition
Yixin Chen’s project [17]	No algorithm is shown in the study	The average value of 10 times prediction	Input: the factors of tablets. Output: dissolution result of a single time point	22 samples and 4 parameters
Jothi G. Kesavan’s project [18]		PROC REG	Input: hardness, friability, and thickness. Output: disintegration time	23 samples and 3 parameters
Uttam MANDAL’s project [19]	ANN using multilayer perceptrons	Higuchi equation	Input: the composition of tablets. Output: dissolution result of a single time point.	13 samples and 3 parameters