aiqbench analysis
This study applies an ensemble of Long Short-Term Memory (LSTM) networks optimized via Particle Swarm Optimization to predict daily river runoff, demonstrating high accuracy ($R^2$ up to 91.42%) for water resource management in the Ajichay River basin. By integrating multiple time intervals and feature importance analysis, the research offers a robust machine learning framework for addressing hydrological stress in data-scarce environments.
A Multi-Timestep LSTM Ensemble regressor for Enhanced Short-Term Runoff Prediction
arXiv:2609.26244v1 Announce Type: new Abstract: Accurately forecasting river runoff is key to managing water resources, controlling floods, and planning agriculture. This study examines the Ajichay River in northwest Iran, a major tributary of Lake Urmia that has experienced increasing water-related stress in recent years. We introduce a daily runoff prediction model based on Long Short-Term Memory (LSTM) networks. The model combines five LSTM units, each trained on different time intervals ranging from 2 to 6 days, to better capture variations in river flow patterns. To improve performance, each model was fine-tuned using Particle Swarm Optimization (PSO), a population-based optimization algorithm. The proposed approach was evaluated on unseen data from 2017-2018 using $R^2$, RMSE, and MSE as performance metrics. The results showed strong predictive accuracy, with $R^2$ values ranging from 74.95% to 91.42%. In addition, multiple feature-importance methods were applied to identify the most influential variables, providing further insight into the factors that drive runoff variations.
Abstract:Accurately forecasting river runoff is key to managing water resources, controlling floods, and planning agriculture. This study examines the Ajichay River in northwest Iran, a major tributary of Lake Urmia that has experienced increasing water-related stress in recent years. We introduce a daily runoff prediction model based on Long Short-Term Memory (LSTM) networks. The model combines five LSTM units, each trained on different time intervals ranging from 2 to 6 days, to better capture variations in river flow patterns. To improve performance, each model was fine-tuned using Particle Swarm Optimization (PSO), a population-based optimization algorithm. The proposed approach was evaluated on unseen data from 2017-2018 using $R^2$, RMSE, and MSE as performance metrics. The results showed strong predictive accuracy, with $R^2$ values ranging from 74.95% to 91.42%. In addition, multiple feature-importance methods were applied to identify the most influential variables, providing further insight into the factors that drive runoff variations.
Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
arXiv:2609.26244 [cs.AI]
(or arXiv:2609.26244v1 [cs.AI] for this version)
https://doi.org/10.48550/arXiv.2609.26244 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 247 words
Related coverage
A Hybrid LSTM-XGBoost Framework for Multi-Horizon Stock Return Prediction Across Diversified Equity Portfolios
arXiv · Sep 14, 2026
Enhanced Evolutionary Multi-Objective Deep Reinforcement Learning for Reliable and Efficient Wireless Rechargeable Sensor Networks
arXiv · Jun 17, 2026
An Auto-Scaling Approach for Serverless Environments Based on a Multi-Expert Consensus Mechanism
arXiv · Jul 20, 2026
An Explainable DistilBERT-BiLSTM-Attention Framework for Binary and Multi-Class Hate Speech Detection
arXiv · Sep 25, 2026