{"id":"arxiv/2610.09996","url":"https://arxiv.org/abs/2610.09996","tags":["stat.AP","math.ST"],"score":0,"title":"Multivariate LSTM Modeling for Wastewater Quality Prediction","domain":"stat.AP","source":"arxiv","authors":["Sophie Laruelle","Innocentia Bénédicta Sossou"],"reviews":0,"subject":"stat.AP","summary":"The proposed approach relies on high-frequency multivariate time series collected at the Seine aval (SIAAP 1 ) treatment plant, including pH, temperature, conductivity, and total suspended solids, as well as exogenous variables related to precipitation and measurements from an upstream plant. The objective is to forecast wastewater quality over a 24-hour horizon. A multivariate Long Short-Term Memory (LSTM) recurrent neural network is implemented to capture complex temporal dependencies and nonlinear patterns in the data. The model is trained on one year of data and validated on four months…","category":"research","comments":0,"disputed":false,"division":"stat","mentions":0,"discussed":0,"published":"2026-10-07","reactions":{}}