Scholarlink

<p>The Eastern Economic Corridor (EEC) initiative focuses on three provinces in Thailand’s eastern region, offering investment incentives to the private sector and funding public transportation infrastructure. This policy drives substantial changes in land use within these areas. In this paper, land use change prediction of the EEC is performed from 2020 to 2040 for Thailand’s public authority. On one hand, land cover of five classes: Urban, Fields, Forest, Industry and Water, is performed using Convolutional Neural Network models. Dataset consists of about 1&#xa0;million tiles representing an area of 0.526&#xa0;km². On the other hand, socio-economic data is analyzed and predicted using five computing methods linear regression, Neural Network, Random Forest, Support-Vector Regression and Extreme Gradient Boosting. The dataset comprises 6 classes representing information from 1960 to 2023: Net Migration, GDP Growth, Unemployment, Inflation, CO<sub>2</sub> Emissions per Capita, and Poverty Ratio. CNN models present an accuracy between 84% (urban class) and 99% (water class). Urban expansion and land reclamation from 2020 to 2040 are projected to significantly reduce agricultural fields, with lesser impacts on forested areas and water bodies. In contrast, urban areas are expected to see substantial growth, particularly inland, while industrial areas remain relatively stable.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Impact of Thailand’s Eastern Economic Corridor (EEC) Policy on Land Use: Prediction Up to 2040 by Combining CNN Land Cover and Machine Learning Methods on Socio-Economic Data

  • 期刊论文

  • 发表日期：

    2025年6月7日

查看全文
  • Damien Gomez,
  • Thibault Lauilhe,
  • Sirikul Hutasavi,
  • Cyrille Schwob,
  • Laurent Mezeix

摘要

The Eastern Economic Corridor (EEC) initiative focuses on three provinces in Thailand’s eastern region, offering investment incentives to the private sector and funding public transportation infrastructure. This policy drives substantial changes in land use within these areas. In this paper, land use change prediction of the EEC is performed from 2020 to 2040 for Thailand’s public authority. On one hand, land cover of five classes: Urban, Fields, Forest, Industry and Water, is performed using Convolutional Neural Network models. Dataset consists of about 1 million tiles representing an area of 0.526 km². On the other hand, socio-economic data is analyzed and predicted using five computing methods linear regression, Neural Network, Random Forest, Support-Vector Regression and Extreme Gradient Boosting. The dataset comprises 6 classes representing information from 1960 to 2023: Net Migration, GDP Growth, Unemployment, Inflation, CO2 Emissions per Capita, and Poverty Ratio. CNN models present an accuracy between 84% (urban class) and 99% (water class). Urban expansion and land reclamation from 2020 to 2040 are projected to significantly reduce agricultural fields, with lesser impacts on forested areas and water bodies. In contrast, urban areas are expected to see substantial growth, particularly inland, while industrial areas remain relatively stable.