Hybrid machine learning modeling of nitrogen removal from wastewater using gas-liquid-solid circulating fluidized bed riser

Keyphrases

• Wastewater 100%
• NO3-N 100%
• Circulating Fluidized Bed Riser 100%
• Machine Learning Modeling 100%
• Nitrogen Removal 100%
• Gas-solid Fluidized Bed 100%
• Hybrid Machine Learning 100%
• Least Squares Support Vector Regression (LSSVR) 75%
• Bayesian Optimization 75%
• Kjeldahl Nitrogen 75%
• Ammonium Ion 50%
• Downer 50%
• Effluent 50%
• High Efficiency 25%
• Multiple Linear Regression 25%
• Prediction Accuracy 25%
• Removal Efficiency 25%
• Computational Efficiency 25%
• Sensitivity Analysis 25%
• Cross-validation Method 25%
• Groundwater Pollution 25%
• Root Mean Square Error 25%
• Noise Level 25%
• Relative Standard Deviation 25%
• Overfitting 25%
• Laboratory Trials 25%
• Comprehensive Evaluation 25%
• K-fold Cross-validation 25%
• Gaussian White Noise 25%
• Eutrophication 25%
• Laboratory Data 25%
• Wastewater Treatment System 25%
• Mean Absolute Percentage Error 25%
• Model Robustness 25%
• Riser 25%
• Circulating Fluidized Bed Bioreactor 25%
• Municipal Wastewater 25%
• Municipal Wastewater Treatment 25%
• Nitrogen N 25%
• Testing Data 25%
• Zero Reference 25%
• Model Hyperparameters 25%
• Essential Variables 25%
• ReliefF Algorithm 25%

Engineering

• Liquid Gas 100%
• Circulating Fluidized Bed 100%
• Downers 100%
• Flow Rate 50%
• Flowrates 50%
• Relative Importance 50%
• Computational Efficiency 50%
• Real Life 50%
• Experimental Observation 50%
• Empirical Model 50%
• Noise Level 50%
• Root Mean Square Error 50%
• Accurate Prediction 50%
• Gaussian White Noise 50%
• Bioreactor 50%
• Municipal Wastewater Treatment 50%
• Reference Line 50%
• Essential Variable 50%
• Cross-Validation Technique 50%

Chemical Engineering

• Learning System 100%
• Fluidized Bed 100%
• Wastewater Treatment 50%
• Biochemical Reactor 50%