![](https://congressarchive.cimne.com/sfmc23/assets/bg_sfmc.jpg)
Comparison of predictive models for influent parameters in the inflow of Water Resource Recovery Facilities
Please login to view abstract download link
This study delves into the use of data exploration and predictive models, including machine learning algorithms such as XGBoost and deep learning neural networks named Long Short-Term Memory (LSTM), to predict the water inflow for Water Resource Recovery Facilities (WRRF). Specifically, it focuses on two influent parameters - inflow water and rain gathered - and utilizes a historical dataset spanning five years to build the predictive model. The model’s validation is done with more recent cases and data from other WRRFs, and it shows promise in helping technicians properly configure WRRF in areas with infrequent but potentially damaging rain events.