Cited 0 time in webofscience Cited 0 time in scopus

Deep Convolutional Neural Networks for estimating PM2.5 concentration levels

Title
Deep Convolutional Neural Networks for estimating PM2.5 concentration levels
Translated Title
PM2.5 농도 추정을 위한 딥 컨볼루션 뉴럴 네트워크
Authors
Kwon, Byung Jun
Advisor(s)
Kwak, Su Ha
Co-Advisor(s)
Moon, Tae Sup
Issue Date
2017
Available Date
2017-08-02
Degree Date
2017. 8
Type
Thesis
Keywords
Deep LearningNeural networksCNNPM2.5
Abstract
Recent, PM2.5 generated by various causes is spreading over a large area. As a result, the technique of estimating the concentration of PM2.5 has become important. Many studies have been conducted to estimate the PM2.5 concentrations. Recently research using learning-based models has been actively conducted. These learning-based models have exceeded the accuracy of existing statistical models. Existing research has a limitation that only data of the monitoring site can be used. In this paper, we propose a method to improve the accuracy of PM2.5 concentration estimation using the information of the surrounding area around the PM2.5 monitoring site. To do this, we study the difference in performance between the 3-dimensional input data and the existing vector data. The algorithm used in this experiment is CNN which can efficiently use the correlation of spatial information. Experimental result show that using large spatial data gives high accuracy of PM2.5 concentration estimation. ⓒ 2017 DGIST
Table Of Contents
Ⅰ. INTRODUCTION 1-- Ⅱ. BACKGROUND 3-- 2.1 Neural Networks 3-- 2.2 Convolutional Neural Networks 5-- Ⅲ. RELATED WORKS 7-- Ⅳ. METERIALS 8-- 4.1 PM2.5 Measurements 8-- 4.2 AOD (Aerosol Optical Depth) data 9-- 4.3 Meteorological fields 10-- 4.4 Land use variables 11-- 4.5 Regional and temporal dummy variables 11-- 4.6 Data integration 11-- Ⅴ. METHOD 13-- 5.1 Data generation 13-- 5.2 Data preprocessing and validation 14-- 5.3 Model structure 15-- Ⅵ. RESULT 18-- Ⅶ. DISCUSSION 24 REFERENCE 25 SUMMARY (Korean) 26
URI
http://dgist.dcollection.net/jsp/common/DcLoOrgPer.jsp?sItemId=000002377545
http://hdl.handle.net/20.500.11750/4103
DOI
10.22677/thesis.2377545
Degree
Master
Department
Information and Communication Engineering
University
DGIST
Files:
Collection:
Emerging Materials ScienceThesesMaster


qrcode mendeley

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

BROWSE