Air Quality Prediction Using Hybrid Methods
Author Information
Author(s): Javier López-Gonzales, Rodrigo Salas, Daira Velandia, Paulo Canas Rodrigues
Primary Institution: Escuela de Posgrado, Universidad Peruana Unión, Lima, Peru
Hypothesis
This study proposes to improve the precision in the prediction of air quality using a hybrid method that integrates singular spectrum analysis and long short-term memory neural networks.
Conclusion
The hybrid method showed high precision in predicting PM10 concentrations and outperformed individual methods.
Supporting Evidence
- The hybrid method achieved better results compared to traditional methods.
- SSA effectively separates signal and noise components, improving prediction accuracy.
- The method is adaptable for real-time monitoring systems.
Takeaway
The researchers created a new way to predict air quality by combining two techniques, making it better at guessing pollution levels.
Methodology
The study used singular spectrum analysis to decompose time series data into trend, seasonal, and noise components, which were then predicted using a long short-term memory neural network.
Limitations
The choice of parameters for singular spectrum analysis can significantly affect the results, and no theoretical solution for optimal parameter selection is provided.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website