A long short-term memory model for forecasting housing prices in Taiwan in the post-epidemic era through big data analytics

IF 5.5 Q1 MANAGEMENT

Asia Pacific Management Review Pub Date : 2024-09-01 DOI:10.1016/j.apmrv.2023.08.002

{"title":"A long short-term memory model for forecasting housing prices in Taiwan in the post-epidemic era through big data analytics","authors":"","doi":"10.1016/j.apmrv.2023.08.002","DOIUrl":null,"url":null,"abstract":"<div><div>This study aims to analyse housing prices in Taiwan in the post-epidemic era, identify the crucial factors influencing them, and develop a suitable method for analysing and forecasting them. This study collects relevant data such as Taiwan's housing price index data from 2002 to 2020 to identify the crucial factors affecting Taiwan's housing prices; this is achieved by constructing a regression model, forecasting Taiwan's housing prices through a constructed long short-term memory (LSTM) model that employs big data analytics, and verifying the efficiency of the proposed models through <em>R-square</em> and root mean square error values. The results indicate that the top 10 factors affecting Taiwan's housing prices are mostly related to mortgage interest rates, suggesting that in Taiwan, the effect on housing prices in the post-epidemic era may be non-significant. This study collects data on Taiwan's housing price for the period from the first quarter of 2002 to the fourth quarter of 2020 to construct an LSTM for forecasting Taiwan's housing prices. The results indicate that the proposed LSTM exhibits good fitness, indicating that the model is suitable for analysing and forecasting housing prices. Given that analysing and forecasting quantity is also crucial in housing market analyses and that this study focuses only on predicting housing prices, future research should explore the simultaneous prediction and analysis of both price and quantity.</div></div>","PeriodicalId":46001,"journal":{"name":"Asia Pacific Management Review","volume":null,"pages":null},"PeriodicalIF":5.5000,"publicationDate":"2024-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S1029313223000623/pdfft?md5=54444bb9bfb81af8514a7f252d65bdf7&pid=1-s2.0-S1029313223000623-main.pdf","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Asia Pacific Management Review","FirstCategoryId":"1085","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S1029313223000623","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"MANAGEMENT","Score":null,"Total":0}

引用次数: 0

Abstract

This study aims to analyse housing prices in Taiwan in the post-epidemic era, identify the crucial factors influencing them, and develop a suitable method for analysing and forecasting them. This study collects relevant data such as Taiwan's housing price index data from 2002 to 2020 to identify the crucial factors affecting Taiwan's housing prices; this is achieved by constructing a regression model, forecasting Taiwan's housing prices through a constructed long short-term memory (LSTM) model that employs big data analytics, and verifying the efficiency of the proposed models through R-square and root mean square error values. The results indicate that the top 10 factors affecting Taiwan's housing prices are mostly related to mortgage interest rates, suggesting that in Taiwan, the effect on housing prices in the post-epidemic era may be non-significant. This study collects data on Taiwan's housing price for the period from the first quarter of 2002 to the fourth quarter of 2020 to construct an LSTM for forecasting Taiwan's housing prices. The results indicate that the proposed LSTM exhibits good fitness, indicating that the model is suitable for analysing and forecasting housing prices. Given that analysing and forecasting quantity is also crucial in housing market analyses and that this study focuses only on predicting housing prices, future research should explore the simultaneous prediction and analysis of both price and quantity.

查看原文本刊更多论文

通过大数据分析预测台湾后疫情时期房价的长短期记忆模型

本研究旨在分析后疫情时代的台湾房价，找出影响房价的关键因素，并制定合适的分析和预测方法。本研究收集了 2002 年至 2020 年台湾住房价格指数数据等相关数据，以确定影响台湾住房价格的关键因素；通过构建回归模型，利用大数据分析构建的长短期记忆（LSTM）模型预测台湾住房价格，并通过 R 平方和均方根误差值验证所提模型的效率。结果表明，影响台湾房价的十大因素大多与房贷利率有关，这表明在台湾，后疫情时代对房价的影响可能并不显著。本研究收集了 2002 年第一季度至 2020 年第四季度的台湾房价数据，构建了预测台湾房价的 LSTM。结果表明，所提出的 LSTM 具有良好的拟合度，表明该模型适用于分析和预测房价。鉴于数量分析和预测在住房市场分析中也至关重要，而本研究仅侧重于预测住房价格，未来的研究应探讨价格和数量的同步预测和分析。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Asia Pacific Management Review MANAGEMENT-

CiteScore

8.00

自引率

4.50%

发文量

期刊介绍： Asia Pacific Management Review (APMR), peer-reviewed and published quarterly, pursues to publish original and high quality research articles and notes that contribute to build empirical and theoretical understanding for concerning strategy and management aspects in business and activities. Meanwhile, we also seek to publish short communications and opinions addressing issues of current concern to managers in regards to within and between the Asia-Pacific region. The covered domains but not limited to, such as accounting, finance, marketing, decision analysis and operation management, human resource management, information management, international business management, logistic and supply chain management, quantitative and research methods, strategic and business management, and tourism management, are suitable for publication in the APMR.