Explainable Graph Neural Networks: An Application to Open Statistics Knowledge Graphs for Estimating House Prices DOI Open Access
Areti Karamanou, Petros Brimos, Evangelos Kalampokis

и другие.

Опубликована: Май 1, 2024

In the rapidly evolving field of real estate economics, prediction house prices continues to be a complex challenge, intricately tied multitude socio-economic factors. However, traditional predictive models have often overlooked spatial interdependencies that play vital role in shaping housing prices. This study applies Graph Neural Networks (GNNs) on Open Statistics Knowledge Graphs model dependencies and predict across Scotland’s 2011 data zones. To this end, integrated statistical indicators are retrieved from official Scottish Government Data portal. The three representative GNN algorithms employed - ChebNet, GCN, GraphSAGE demonstrate higher accuracy than models, including tabular-based XGBoost simple Multi-Layer Perceptron (MLP). addition, local global explainability increase transparency trust predictions made by most accurate GraphSAGE. feature importance is determined logistic regression surrogate while local, region-level understanding achieved through use GNNExplainer. Explainaibility results compared with those previous work applied machine learning algorithm SHapley Additive exPlanations (SHAP) framework same dataset. Interestingly, both SHAP approach underscored Comparative Illness Factor, health indicator, ratio detached dwellings as crucial features explainability. case explanations, methods showed similar results, provided richer, more comprehensive for two specific

Язык: Английский

Unraveling the anchoring effect of seller’s show on buyer’s show to enhance review helpfulness prediction: A multi-granularity attention network model with multimodal information DOI
Feifei Wang,

Z. Zhang,

Jie Song

и другие.

Electronic Commerce Research and Applications, Год журнала: 2025, Номер unknown, С. 101484 - 101484

Опубликована: Янв. 1, 2025

Язык: Английский

Процитировано

0

Comparative Analysis of Advanced Models for Predicting Housing Prices: A Review DOI Creative Commons

Inmaculada Moreno-Foronda,

María Teresa Sánchez Martínez, Montserrat Pareja‐Eastaway

и другие.

Urban Science, Год журнала: 2025, Номер 9(2), С. 32 - 32

Опубликована: Янв. 31, 2025

Understanding the determinants of housing price movements is an ongoing subject debate. Estimating these becomes a valuable tool for predicting trends and mitigating risks market volatility. This article presents systematic review analyzing studies that compare various machine learning (ML) tools with hedonic regression, aiming to assess whether real estate predictions based on mathematical techniques artificial intelligence enhance accuracy models used valuing residential properties. ML (neural networks, decision trees, random forests, among others) provide high predictive capacity greater explanatory power due better fit their statistical measures. However, regression models, while less precise, are more robust, as they can identify attributes most influence levels. These include property’s location, its internal features, distance from property city centers.

Язык: Английский

Процитировано

0

Forecasting second-hand house prices in China using the GA-PSO-BP neural network model DOI Creative Commons
Jining Wang, Haoran Ji, Lei Wang

и другие.

PLoS ONE, Год журнала: 2025, Номер 20(5), С. e0322821 - e0322821

Опубликована: Май 7, 2025

While the traditional genetic algorithms are capable of forecasting house prices, they often suffer from premature convergence, which adversely affects reliability forecasts. To address this issue, research employs a genetic-particle swarm optimization (GA-PSO) algorithm and develops GA-PSO-BP neural network model through integration BP network. Building upon foundation, study considers several pivotal factors affecting housing prices dataset comprising 1,824 transactions second-hand homes 2023 to 2024, gathered Lianjia.com, forecast in China. This work shows that demonstrates exceptional performance when dealing with complex high-dimensional data, significantly minimizing errors. The test set achieved an RMSE 0.786 MAPE 8.9%. Its effectiveness houses notably surpasses optimized by single algorithm. provides more accurate forecasts rapidly growing urban areas such as Guangzhou, thus providing essential insights for investors contemplating real estate investment.

Язык: Английский

Процитировано

0

The soft computing based model of investors’ condition and cognition on a real estate market DOI
Małgorzata Renigier‐Biłozor, Aneta Chmielewska, Ewelina Kamasz

и другие.

Land Use Policy, Год журнала: 2024, Номер 141, С. 107150 - 107150

Опубликована: Март 25, 2024

Язык: Английский

Процитировано

1

Explainable Graph Neural Networks: An Application to Open Statistics Knowledge Graphs for Estimating House Prices DOI Creative Commons
Areti Karamanou, Petros Brimos, Evangelos Kalampokis

и другие.

Technologies, Год журнала: 2024, Номер 12(8), С. 128 - 128

Опубликована: Авг. 6, 2024

In the rapidly evolving field of real estate economics, prediction house prices continues to be a complex challenge, intricately tied multitude socio-economic factors. Traditional predictive models often overlook spatial interdependencies that significantly influence housing prices. The objective this study is leverage Graph Neural Networks (GNNs) on open statistics knowledge graphs model these dependencies and predict across Scotland’s 2011 data zones. methodology involves retrieving integrated statistical indicators from official Scottish Open Government Data portal applying three representative GNN algorithms: ChebNet, GCN, GraphSAGE. These GNNs are compared against traditional models, including tabular-based XGBoost simple Multi-Layer Perceptron (MLP), demonstrating superior accuracy. Innovative contributions include use in economics application local global explainability techniques enhance transparency trust predictions. feature importance determined by logistic regression surrogate while local, region-level understanding predictions achieved through GNNExplainer. Explainability results with those previous work applied machine learning algorithm SHapley Additive exPlanations (SHAP) framework same dataset. Interestingly, both SHAP approach underscored comparative illness factor, health indicator, ratio detached dwellings as most crucial features explainability. case explanations, methods showed similar results, provided richer, more comprehensive for two specific

Язык: Английский

Процитировано

1

Enhancing Housing Price Prediction Using AI and Machine Learning: A Stacked Regression Meta-Modeling Approach DOI

Malgi Nikitha Vivekananda,

Prashant Ashok Shidlyali

Опубликована: Окт. 3, 2024

Язык: Английский

Процитировано

1

The role of marketing in the dynamics of real estate leasing in Peru: findings, challenges and solutions DOI Creative Commons
Jenrry Anibal Flores Vasquez, Marco Antonio Hernández Muñóz, Victor Hugo Puican Rodríguez

и другие.

Journal of Law and Sustainable Development, Год журнала: 2023, Номер 11(11), С. e1133 - e1133

Опубликована: Ноя. 14, 2023

Purpose: The main objective of this study is to quantify the impact marketing strategies on real estate leasing in Peruvian context. Theoretical framework: An exhaustive review academic literature was carried out gain an in-depth knowledge existing paradigms related and phenomenon. Design/Methodology/Approach: A quantitative, descriptive-explanatory methodology chosen. structured questionnaire administered a representative sample 30 tenants. Results: data collected evidenced notable correlation between tactics lease rates, with significant p-value (less than 0.05). Also, Spearman's Rho Kendall's Tau_b coefficients 0.678 0.632, respectively, were found. It observed that approximately half contracts analyzed are not duly registered SUNARP, there lack detailed information tenant profile about one third developments. Practical social implications: registration generates environment legal vulnerability, increasing risk conflicts parties involved. absence tenant's may hinder proper selection Emphasis placed proposal establish effective conflict resolution mechanisms imperative need for transparency fee structures, seeking strengthen fiduciary relationship landlords Originality/value: This provides innovative view sector. empirical quantitative evidence current contractual practices presentation properties market. highlights urgent refine consolidate transparent reliable market Peru.

Язык: Английский

Процитировано

1

The Role of Marketing in The Dynamics of Real Estate Leasing in Peru: Findings, Challenges and Solutions DOI Creative Commons
Jenrry Anibal Flores Vasquez, Marco Antonio Hernández Muñóz, Victor Hugo Puican Rodríguez

и другие.

Revista de Gestão Social e Ambiental, Год журнала: 2024, Номер 18(1), С. e04918 - e04918

Опубликована: Фев. 20, 2024

Purpose: The main objective of this study is to quantify the impact marketing strategies on real estate leasing in Peruvian context. Theoretical framework: An exhaustive review academic literature was carried out gain an in-depth knowledge existing paradigms related and phenomenon. Design/Methodology/Approach: A quantitative, descriptive-explanatory methodology chosen. structured questionnaire administered a representative sample 30 tenants. Results: data collected evidenced notable correlation between tactics lease rates, with significant p-value (less than 0.05). Also, Spearman's Rho Kendall's Tau_b coefficients 0.678 0.632, respectively, were found. It observed that approximately half contracts analyzed are not duly registered SUNARP, there lack detailed information tenant profile about one third developments. Practical social implications: registration generates environment legal vulnerability, increasing risk conflicts parties involved. absence tenant's may hinder proper selection Emphasis placed proposal establish effective conflict resolution mechanisms imperative need for transparency fee structures, seeking strengthen fiduciary relationship landlords Originality/value: This provides innovative view sector. empirical quantitative evidence current contractual practices presentation properties market. highlights urgent refine consolidate transparent reliable market Peru.

Язык: Английский

Процитировано

0

Explainable Graph Neural Networks: An Application to Open Statistics Knowledge Graphs for Estimating House Prices DOI Open Access
Areti Karamanou, Petros Brimos, Evangelos Kalampokis

и другие.

Опубликована: Май 1, 2024

In the rapidly evolving field of real estate economics, prediction house prices continues to be a complex challenge, intricately tied multitude socio-economic factors. However, traditional predictive models have often overlooked spatial interdependencies that play vital role in shaping housing prices. This study applies Graph Neural Networks (GNNs) on Open Statistics Knowledge Graphs model dependencies and predict across Scotland’s 2011 data zones. To this end, integrated statistical indicators are retrieved from official Scottish Government Data portal. The three representative GNN algorithms employed - ChebNet, GCN, GraphSAGE demonstrate higher accuracy than models, including tabular-based XGBoost simple Multi-Layer Perceptron (MLP). addition, local global explainability increase transparency trust predictions made by most accurate GraphSAGE. feature importance is determined logistic regression surrogate while local, region-level understanding achieved through use GNNExplainer. Explainaibility results compared with those previous work applied machine learning algorithm SHapley Additive exPlanations (SHAP) framework same dataset. Interestingly, both SHAP approach underscored Comparative Illness Factor, health indicator, ratio detached dwellings as crucial features explainability. case explanations, methods showed similar results, provided richer, more comprehensive for two specific

Язык: Английский

Процитировано

0