A Study on the Prediction on the Emergence of Urban Hotspot Regions Based on Machine Learning Algorithms

DOI IR (HANDLE) Web Site Open Access
  • REN Yujie
    Graduate School of Human-Environment Studies, Kyushu University : Doctoral Program
  • ZHAO Shichen
    Faculty of Human-Environmet Studies, Kyushu University
  • DU Mengge
    Graduate School of Human-Environment Studies, Kyushu University : Doctoral Program

Bibliographic Information

Other Title
  • 機械学習方法に基づく都市ホットスポットの発生予測に関する研究

Search this article

Description

This study predicts the emergence state, date, time period and function of urban hotspots and also compares the performance of different data sources (numerical data, digital map vector data, remote sensing raster map data) combined with different algorithms (regression model, random forest and convolution neural network) in urban hotspot prediction. The results illustrate that the combinations of different data sources and algorithms have different prediction accuracy in various scenarios. This study provides guidance for urban hotspot prediction under different scenarios and objectives.

Journal

  • 都市・建築学研究

    都市・建築学研究 41 15-23, 2022-01-15

    Faculty of Human-Environment Studies, Kyushu University

Keywords

Details 詳細情報について

Report a problem

Back to top