A Long Short-Term Memory Neural Network Model for Predicting Air Pollution Index Based on Popular Learning

EasyChair Preprint 2586

Abstract

With the acceleration of industrialization and modernization, the problem of air pollution has become more and more prominent, causing serious impact on people's production and life. Therefore, it is of great practical significance and social value to realize the prediction of air quality index. This article takes the analysis of Tianjin air quality data and meteorological data from 2017 to 2019 as an example.Firstly, random forest interpolation was used to fill in missing values in the data reasonably. Secondly, in the framework of Tensorlow deep learning, Locally Linear Embedding (LLE) was used to screen multivariate data to reduce data dimensions and realize feature selection.Finally, a prediction model of the air quality index was established by using the Long Short-Term Memory (LSTM) neural network based on the data after dimension reduction. The experimental results show that the method has obvious effects in terms of dimensionality reduction and exponential prediction accuracy compared with Principal Component Analysis (PCA) and Back Propagation (BP).

Keyphrases: Air quality prediction, LLE, LSTM, deep learning

@booklet{EasyChair:2586,
  author    = {Hong Fang and Yibo Feng and Lan Zhang and Ming Su and Hairong Yang},
  title     = {A Long Short-Term Memory Neural Network Model for Predicting Air Pollution Index Based on Popular Learning},
  howpublished = {EasyChair Preprint 2586},
  year      = {EasyChair, 2020}}