Preparation of Dietary Fiber Modified by Taro Peel and Its NO2- Adsorption Capacity Prediction Based on Algorithm Optimization Extreme Learning Machine
Journal Title: Chinese Journal of Inorganic Analytical Chemistry - Year 2025, Vol 15, Issue 6
Abstract
Based on the response surface methodology,all experimental data,including process parameters and NO2- adsorption capacities were collected. Following data preprocessing,suitable input variables (solid-to-liquid ratio,hydrochloric acid concentration,reaction temperature,and reaction time) were selected. An initial Extreme Learning Machine (ELM) model was constructed using the training dataset. Genetic Algorithm (GA),Particle Swarm Optimization (PSO),Sparrow Search Algorithm (SSA),Grey Wolf Optimizer (GWO),and Seagull Optimization Algorithm (SOA) were employed to optimize the ELM. The optimized ELM models were trained using the training dataset. The performance of models was evaluated using a testing dataset. Models was assessed by various performance metrics. The results showed that five optimized ELM models outperformed initial ELM model in all performance metrics. Among the five optimization algorithms,SSA-ELM model exhibited the most significant performance. MAE (Mean Absolute Error),MSE (Mean Squared Error),RMSE (Root Mean Squared Error),and MAPE (Mean Absolute Percentage Error) values were 0.023 498,0.000 739 1,0.027 186,and 0.037267%. It was the lowest among all tested models. In terms of the coefficient of determination (R2),the original ELM model had an R2 value of 0.013 291. The R2 values for the GA-ELM,PSO-ELM,SSA-ELM,GWO-ELM,and SOA-ELM models were 0.867 09,0.980 16,0.999 71,0.999 98,and 0.999 69,respectively. This indicated that the five optimized ELM models possess higher fitting accuracy,better generalization ability,and stability. R2 was a significant improvement values to compare with the original ELM model. The optimized ELM models provided a rapid and accurate predictive tool for estimating NO2- adsorption capacities of modified taro peel dietary fiber under varying processing conditions. This not only reduces experimental costs and time but also enhances production efficiency and product quality,thereby offering a reliable and robust solution for practical applications in the field.
Authors and Affiliations
Zhonghui DENG, Wei XIE
Preparation of Dietary Fiber Modified by Taro Peel and Its NO2- Adsorption Capacity Prediction Based on Algorithm Optimization Extreme Learning Machine
Based on the response surface methodology,all experimental data,including process parameters and NO2- adsorption capacities were collected. Following data preprocessing,suitable input variables (solid-to-liquid ratio...
Identification of Degradability of Plastic Films Based on Stable Carbon Isotope Ratios
To explore the relationship among carbon content,stable carbon isotope ratio (δ13C),and the degradability of plastic films,89 commercially available plastic film products were collected from diverse platforms. These p...
Rapid Quantification of Heavy Metals in Oil Sludge Based on Laser-Induced Breakdown Spectroscopy Combined with Random Forest Algorithm
The quantitative analysis of heavy metal elements in oily sludge is of great significance for the assessment of soil environmental pollution in petroleum and petrochemical industrial areas. In this study,the laser-induc...
Rapid Determination of Heavy Metals in Industrial Wastewater by X-ray Fluorescence Spectrometry with Filter Membrane Enrichment Carriers
X-ray fluorescence analysis is a widely used method for measuring heavy metals,characterized by simple operation,rapid analysis,and non-destructive testing. However,its applications are mainly focused on the analysis...
Determination of Ag in Geochemical Samples by Graphite Furnace Atomic Absorption Spectrometry with Triacidic Decomposition-Thiourea Complexing
The determination of Ag content in geochemical samples typically involves treatment by aqua regia decomposition-hydrochloric acid or aqua regia extraction. However,aqua regia fails to completely dissolve the samples,an...