Scholarlink

In the landscape of housing market research, there exists a predominant focus on analytical methodologies and economic influencers. However, housing markets stand as pivotal pillars shaping stakeholder contributions to macroeconomic growth. This study explores extensive UK housing data spanning several years, employing an integrated statistical and machine learning techniques approach. By facilitating dynamic updates on comprehensive fluctuations and offering tailored forecasts while quantifying reliability, this research aims to streamline the complexities of market dynamics by harnessing the power of different statistical and machine learning methodologies, data analysis, and visualisation techniques. The project dissects significant factors impacting the UK housing market from 1995 to 2023. Additionally, it seeks to extend its foresight to 2033. This research novelty resides in the proposed integrated framework combining macroeconomic interpretation with Machine Learning in R and interactive Power BI tool with statistical analysis to forecast sales volume. By leveraging and comparing methods like ARIMA and the combination of STL with ETS, alongside neural network models like LSTM, this framework aims to offer valuable market insights and predictions. The results of RMSE and MSE using ARIMA and LSTM models validate the proposed data modelling and analysis in our specific research context. These insights are crucial for stakeholders like policymakers, aiding informed decision-making and enhancing understanding of the UK housing market dynamics.

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

A Hybrid Machine Learning and Statistical Analysis Approach for Sales Forecasting and Patterns Discovering in the UK Housing Market

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Cheima Ali Bensaad,
  • Dileep Singh

摘要

In the landscape of housing market research, there exists a predominant focus on analytical methodologies and economic influencers. However, housing markets stand as pivotal pillars shaping stakeholder contributions to macroeconomic growth. This study explores extensive UK housing data spanning several years, employing an integrated statistical and machine learning techniques approach. By facilitating dynamic updates on comprehensive fluctuations and offering tailored forecasts while quantifying reliability, this research aims to streamline the complexities of market dynamics by harnessing the power of different statistical and machine learning methodologies, data analysis, and visualisation techniques. The project dissects significant factors impacting the UK housing market from 1995 to 2023. Additionally, it seeks to extend its foresight to 2033. This research novelty resides in the proposed integrated framework combining macroeconomic interpretation with Machine Learning in R and interactive Power BI tool with statistical analysis to forecast sales volume. By leveraging and comparing methods like ARIMA and the combination of STL with ETS, alongside neural network models like LSTM, this framework aims to offer valuable market insights and predictions. The results of RMSE and MSE using ARIMA and LSTM models validate the proposed data modelling and analysis in our specific research context. These insights are crucial for stakeholders like policymakers, aiding informed decision-making and enhancing understanding of the UK housing market dynamics.