Removal efficiency optimization of Pb2+ in a nanofiltration process by MLP-ANN and RSM

Modeling and optimization of malondialdehyde (MDA) absorbance behavior through response surface methodology (RSM) and artificial intelligence network (AIN): An endeavor to estimate lipid peroxidation by determination of MDA

“…Both time‐ and cost‐effectiveness data‐driven algorithms for finding solutions are among the elements contributing to the theory, design, application, and development of biological and linguistical computational paradigms. In machine learning, the patterns are studied in datasets according to parameters such as gene expression and clinical information within a multitude of cases and have been utilized particularly in biomarker discovery research studies 18,45–47 …”

“…More importantly, when the number of variables increases and too many complicated interactions obtain, describing the object functions cannot be conducted by RSM accurately. However, ANN has been succeeded in overcoming the RSM limitations in terms of data analysis with noise effectively, mimicking biological neural networks, efficient performance in comparison with statistical/regression‐based models as well as generating accurate prediction and modeling highly nonlinear and complicated biological processes 45,46,48 …”

“…To design and optimize the influential variables in the process, RSM, as a statistical and mathematical tool, was introduced to examine the effects of different factors 44–46,57 . The number of trials is reduced and recognized because of parameters affecting the process of the removal procedure.…”

“…This methodology has successfully optimized the process parameters, such as adsorbent dose, pH, metal ion concentration, reaction time etc., that are applied for the bio‐sorption of metals and dyes. RSM can also perform optimal design to reduce the cost of experimental works, which can be represented graphically in various plots to illustrate the shape of the response surface 45 …”

“…In machine learning, the patterns are studied in datasets according to parameters such as gene expression and clinical information within a multitude of cases and have been utilized particularly in biomarker discovery research studies. 18,[45][46][47] The optimization process is considered a simple, common one-factor-at-a-time (OFAT) strategy. 18 However, the critical disadvantages of this approach are the time consumed and inaccuracy, with less than optimum expectations.…”