Air quality forecasting using neural networks

Abstract

In this thesis project, a special type of neural network: Extreme Learning Machine (ELM) is implemented to predict the air quality based on the air quality time series itself and the external meteorological records. A regularized version of ELM with linear components is chosen to be the main model for prediction. To take full advantage of this model, its hyper-parameters are studied and optimized. Then a set of variables is selected (or constructed) to maximize the performance of ELM, where two different variable selection methods (i.e. wrapper and filtering methods) are evaluated. The wrapper method ELM-based forward selection is chosen for the variable selection. Meanwhile, a feature extraction method (Principal Component Analysis) is implemented in the hope of reducing the candidate meteorological variables for feature selection, which proves to be helpful. At last, with all the parameters being properly optimized, ELM is used for the prediction and generates satisfying results.

Air Quality Prediction in Smart Cities Using Machine Learning Technologies Based on Sensor Data: A Review

“…Air Quality Forecasting using Neural Networks [35]: Zhao et al suggested to apply extreme learning machine-based approach to forecast air quality. The case study was Helsinki, and the data included hourly air quality data (nitric oxide (NO), O 3 , PM 10 , PM 2.5 ) and meteorological data (relative humidity, pressure, temperature, and wind).…”

Real-Time Profiling of Fine-Grained Air Quality Index Distribution Using UAV Sensing

“…As it has been proved that a three-layer neural network can compute any arbitrary function [32]- [34], NN is able to present the complicated changes in fine-grained scenario. However, without considering the physical characteristics of AQI, the NN model may overfit and perform worse on the test data than on the training data [18].…”

“…However, they mainly focus on 2D area, and can hardly produce a 3D fine-grained map. Neural networks (NN) are also used for forecasting on the AQI distribution [18]- [21]. However, its performance in fine-gained area is not satisfied without considering the physical characteristic of real AQI distribution.…”

“…The neural network model, especially multilayer perceptron (MLP), has been wildly adopted to do estimation for air quality [18]- [21]. They usually train models by using a huge amount of data to achieve decent performance.…”

Accuracy improvement in air‐quality forecasting using regressor combination with missing data imputation

“…Kumar and Goyal 10 applied the principal component regression method to forecast short-term daily air-quality index for four seasons in India. Zhao et al 11 proposed a regularized version of an extreme learning machine for air-quality forecasting. Zhou et al 12 trained a multi-output support vector machine (M-SVM) model by using multi-task learning algorithm to capture nonlinear relations and share the correlation information across related tasks.…”