Advancing the Comprehension of Iron Ions with Different Valences in Anammox-Based Processes: Insight into the Response and Mechanism Based on Prediction and Classification Machine Learning Models

• DOI: 10.1021/acsestwater.3c00543
• Publication Date: 2023-10-10T05:18:04Z
• Authors: Zhicheng Jiang; Xinxin Xu; Yuhang He; Ming Zeng; Meng Zhang; Wei Liu; Nan Wu
• Keywords: Anammox

The responses of anammox to Fe2+ and Fe3+ have been widely discussed; however, the critical concentration is not certain due different culture environments. A reliable applicable predictive classification system based on anammox-based with iron-containing wastewater needs be established, which can surmount independence between experiments combined statistical analysis elucidate mechanism effect concentrations. results confirmed that 5 mg/L iron promoted nitrogen removal process, while higher concentrations inhibited a ratio above 80%. Moreover, Spearman correlation proved showed more obvious rate than Fe3+. To precisely predict performance anammox, Support Vector Machine, XGBoost, Random Forest (RF) were compared, RF model was as preferable (R2 = 0.99). According model, influence ions could successfully traced back valent states an accuracy 97.7%. Furthermore, process in under ion stress proposed.