Download PDFOpen PDF in browser
ZH
Switch back to the title and the abstract
in Chinese

Prediction of Air Quality Index Based on LSTM Model: A Case Study on Delhi and Houston

EasyChair Preprint 1214

7 pagesDate: June 20, 2019

Abstract

Air Quality Index (AQI) prediction is one of the hot topics in Air Quality research, which is helpful to evaluate the impact of urban air pollutants on human health.In the past ten years, people have learned that air pollution is a serious problem, Air Quality Index is a number, based on the synthetical effects of the main air pollutant concentration, by government agencies to evaluate the air quality at different sites, is also used in many of the world metropolis local and regional air quality management.Traditional prediction methods can only analyze small amount of air quality data, which leads to low prediction accuracy, slow speed, low efficiency and easy data fitting. In order to meet the requirements of users for real-time data processing.In this paper, a recursive neural network model based on Long short-term memory (LSTM) is proposed. Because it can effectively utilize the long-distance dependent information in sequence data, it is very suitable for the prediction of air quality index.In this paper, the air quality data of Delhi and Houston in recent years are combined for regression fitting to predict the future air quality index. This model uses MAPE, RMSE, R, IA and MAE to process the data, and is compared with MLR(BGD), MLR(SGD), MLR(MBGD) and regression model (SVR).Experimental results show that LSTM neural network can accurately predict AQI.

Keyphrases: LSTM network, air quality, index forecast

BibTeX entry
BibTeX does not have the right entry for preprints. This is a hack for producing the correct reference:
@booklet{EasyChair:1214,
  author    = {Dongwen Zhang and Qi Zhao and Yunfeng Xu},
  title     = {Prediction of Air Quality Index Based on LSTM Model: A Case Study on Delhi and Houston},
  howpublished = {EasyChair Preprint 1214},
  year      = {EasyChair, 2019}}
Download PDFOpen PDF in browser