Housing Dataset Research Articles

Marginalized LASSO in the low-dimensional difference-based partially linear model for variable selection

The difference-based partially linear model is an appropriate regression model when both linear and nonlinear predictors are present in the data. However, when we want to optimize the weights using the difference-based method, the problem of variable selection can be difficult since low-variance predictors present a challenge. Therefore, this study aims to establish a novel methodology based on marginal theory to tackle such mixed relationships successfully, emphasizing variable selection in low dimensions. We suggest using a marginalized LASSO estimator with a penalty term that is not as severe and related to the difference order. As part of our numerical analysis of small sample performance, we undertake comprehensive simulation experiments to numerically demonstrate the strength of our proposed technique in estimation and prediction compared to the LASSO under a low-dimensional setup. This is done so that we can numerically demonstrate the strength of our proposed method in estimation and prediction. The bootstrapped method is utilized to evaluate how well our proposed prediction method performs when examining the King House dataset.

High School Sports Houses as Identity Primers: Constructing Queensland and Australian Identities

ABSTRACT Sport plays an important role in the conception of Australian national identity. Much of the scholarly work on this topic has concentrated on professional, high-profile sports and athletes. In this article, we draw attention to a highly symbolic aspect of popular sporting culture—school sports houses—and the prominent historical figures and motifs that they project through their team names, mascots and colours. School sports houses are potent examples of banal nationalism, whereby elements of Australianness are reinforced through the tradition of school athletics and swimming carnivals. Here, we highlight the overlap between the school-based agenda of encouraging group identity and the broader notion of the nation as an imagined community. Using a comprehensive dataset of Queensland high school sport houses in 1969 and 2023, we observe several trends, such as the persistence of houses named after colonial explorers and the disparity between houses with women’s and men’s namesakes. We consider the role of school sports houses as “identity primers”, or scripts, that encourage schoolchildren to identify with specific historic figures, landforms, animal totems, character traits and other qualities as indicators of the region and nation in which they live.

Diminishing benefits of thermal mass in Iranian climate: Present and future scenarios

Thermal mass, a pivotal element in a building's performance, functions as an indoor thermal buffer. While literature underscores its advantages, the enduring impact of thermal mass amid climate change remains uncertain. This study methodically assesses thermal mass effects in 21 Iranian cities across contemporary and future climates, juxtaposing heavyweight and lightweight constructions. The EPSAP algorithm, a generative building design method, created a dataset of two-story single-family houses. Cooling and heating demands were evaluated in EnergyPlus, accounting for current and future system design efficiencies. Future climates were simulated using EC-Earth3 model estimations for the SSP5-8.5 scenario in 2050 and 2080 timeframes. The findings reveal that the energy efficiency advantage of heavyweight over lightweight buildings will diminish by up to 0.60 kW·h·m−2 in 2050 (40 % less than the present-day climate difference between constructions) and 0.93 kW·h·m−2 in 2080 (63 %) for cities in central and southern regions. The performance differences between constructions will sometimes be null, making thermal mass negligible. Conversely, only three cities in Northern Iran exhibit an opposing trend for mid to very-high thermal transmittances. Regarding building geometry, heavyweight construction correlates strongly with indexes related to building compactness, while lightweight construction aligns more with glazing-related indexes. However, as climates warm or we move towards warmer regions, discernible differences between lightweight and heavyweight constructions vanish for both shape- and glazing-related indexes. In conclusion, although the use of thermal mass will be less effective, building design professionals will have greater latitude for innovative construction and design solutions.

Parameter Estimation in Spatial Autoregressive Models with Missing Data and Measurement Errors

This study addresses the problem of parameter estimation in spatial autoregressive models with missing data and measurement errors in covariates. Specifically, a corrected likelihood estimation approach is employed to rectify the bias in the log-maximum likelihood function induced by measurement errors. Additionally, a combination of inverse probability weighting (IPW) and mean imputation is utilized to mitigate the bias caused by missing data. Under several mild conditions, it is demonstrated that the proposed estimators are consistent and possess oracle properties. The efficacy of the proposed parameter estimation process is assessed through Monte Carlo simulation studies. Finally, the applicability of the proposed method is further substantiated using the Boston Housing Dataset.

Integrating Dropout and Kullback-Leibler Regularization in Bayesian Neural Networks for improved uncertainty estimation in Regression

The objective of the study is to enhance uncertainty prediction in regression problems by introducing a revolutionary Bayesian Neural Network (BNN) model. Experimental results reveal significant improvements in uncertainty prediction and point forecasts with the integrated BNN model compared to the plain BNN. Performance metrics, including mean squared error (MSE), mean absolute error (MAE), and R-squared (R²), demonstrate superior results for the proposed BNN. The experimental results show that for plain BNN, MSE is 87.3, MAE is 6.62 and R2 is −0.0492, whereas for the proposed BNN model MSE was found to be 44.64, MAE is 4.4 and R2 is 0.46. This research brings a fresh approach to Bayesian Neural Networks by incorporating both dropout and KL regularization techniques, resulting in a powerful tool for handling regression tasks with certainty. By combining these techniques, the study enhances model stability, avoids overfitting, and achieves more reliable uncertainty estimation. This study adds to our knowledge of uncertainty-aware machine learning models and offers a valuable solution for accurately assessing uncertainty in various applications.•The innovative BNN model merges the power of Bayesian principles with the effectiveness of dropout and KL regularization.•To test and refine our model, the study utilizes the Boston Housing dataset for both training and evaluation purposes.

The spatial structure of housing affordability and the impact of public infrastructure

ABSTRACT We develop a parsimonious agent-level model of the housing market that reproduces the spatial price structure of London at the level of each property. Prices emerge as agents make utility-driven choices on properties with specific attributes in decentralised market interactions. This specification facilitates matching individual households to specific properties, something not possible in most publicly available large-scale housing datasets. Because location is a feature of properties, we estimate the impact of the new Elizabeth Line that connects the city from east to west. Then, by combining household and property variables, we estimate its impact on affordability. Changes in affordability allow to foresee potential demographic changes across space. Such information is extremely valuable not only to households and developers, but also to planning authorities who wish to anticipate the future demand for services and test the viability of different local tax regimes.

An Optimal House Price Prediction Algorithm: XGBoost

An accurate prediction of house prices is a fundamental requirement for various sectors, including real estate and mortgage lending. It is widely recognized that a property’s value is not solely determined by its physical attributes but is significantly influenced by its surrounding neighborhood. Meeting the diverse housing needs of individuals while balancing budget constraints is a primary concern for real estate developers. To this end, we addressed the house price prediction problem as a regression task and thus employed various machine learning (ML) techniques capable of expressing the significance of independent variables. We made use of the housing dataset of Ames City in Iowa, USA to compare XGBoost, support vector regressor, random forest regressor, multilayer perceptron, and multiple linear regression algorithms for house price prediction. Afterwards, we identified the key factors that influence housing costs. Our results show that XGBoost is the best performing model for house price prediction. Our findings present valuable insights and tools for stakeholders, facilitating more accurate property price estimates and, in turn, enabling more informed decision making to meet the housing needs of diverse populations while considering budget constraints.

Robust Variable Selection with Exponential Squared Loss for the Spatial Error Model

With the widespread application of spatial data in fields like econometrics and geographic information science, the methods to enhance the robustness of spatial econometric model estimation and variable selection have become a central focus of research. In the context of the spatial error model (SEM), this paper introduces a variable selection method based on exponential square loss and the adaptive lasso penalty. Due to the non-convex and non-differentiable nature of this proposed method, convex programming is not applicable for its solution. We develop a block coordinate descent algorithm, decompose the exponential square component into the difference of two convex functions, and utilize the CCCP algorithm in combination with parabolic interpolation for optimizing problem-solving. Numerical simulations demonstrate that neglecting the spatial effects of error terms can lead to reduced accuracy in selecting zero coefficients in SEM. The proposed method demonstrates robustness even when noise is present in the observed values and when the spatial weights matrix is inaccurate. Finally, we apply the model to the Boston housing dataset.

Enhancing Neural Network with Particle Swarm: A House Price Prediction Case

This study delves into the utilization of the Particle Swarm Optimization (PSO) algorithm to discover the best parameters of Long Short-Term Memory (LSTM) model with the aim of housing price prediction. Employing the Ames Housing dataset, our investigation underscores the remarkable efficacy of PSO in mitigating prediction errors within the LSTM framework, thereby enhancing overall predictive performance. The symbiotic integration of PSO with LSTM facilitates the discernment of intricate data patterns, particularly in scenarios involving high-dimensional and nonlinear optimization conundrums. The findings illuminate PSO's ability to expediently traverse and identify global optima. The outcomes unequivocally underscore the PSO-LSTM model's superior attributes, characterized by accelerated convergence, diminished errors, and exceptional proficiency. In summation, this investigation underscores PSO's promise as a potent technique for optimizing machine learning models, notably exemplified by its integration with the LSTM framework.

NIA Genetics of Alzheimer’s Disease Data Storage Site (NIAGADS): 2023 Update

AbstractBackgroundNIAGADS is a national genomics data repository that facilitates access of genotypic and sequencing data to qualified investigators for the study of the genetics of Alzheimer’s disease (AD) and related neurological diseases. Collaborations with large consortia and centers such as the Alzheimer’s Disease Genetics Consortium (ADGC), Cohorts for Heart and Aging Research in Genomic Epidemiology (CHARGE) Consortium, the Alzheimer’s Disease Sequencing Project (ADSP), and the Genome Center for Alzheimer’s Disease (GCAD) allow NIAGADS to lead the effort in managing large AD datasets that can be easily accessed and fully utilized by the research community.MethodNIAGADS is supported by National Institute on Aging (NIA) under a cooperative agreement. All data derived from NIA funded AD genetics studies are expected to be deposited in NIAGADS or another NIA approved site. NIAGADS manages a Data Sharing Service (DSS) that facilitates the deposition and sharing of genomic data and association results with approved users in the neurodegenerative research community. In addition, researchers are able to freely use the NIAGADS Alzheimer’s Genomics Database (www.niagads.org/genomics/) to search annotation resources that link published AD studies to AD‐relevant sequence features and genome‐wide annotations.ResultAs of January 2023, NIAGADS houses 93 datasets comprised of >149,000 samples including GWAS, sequencing, gene expression, annotations, deep phenotypes, and summary statistics. Qualified investigators can retrieve ADSP sequencing data with ease and flexibility through the NIAGADS DSS. To date, the ADSP and other contributing studies have completed whole exome sequencing (WES) of 20,503 samples and whole genome sequencing (WGS) of 36,361 samples. Raw WES and WGS files, quality controlled VCF files, and phenotype data files are available via qualified access. The next round of sequencing currently underway will generate around 30,000 additional genomes to be released in late 2023.ConclusionNIAGADS is a rich resource for AD researchers, with the goal of facilitating advances in Alzheimer’s genetics research. By housing datasets from many projects and institutions, NIAGADS enables AD researchers to meet their research goals more efficiently. Datasets, guidelines, and features are available on our website at https://www.niagads.org.

House Prices Prediction Advanced Regression Techniques

In recent years, with the development of the real estate industry, housing prices have continued to rise. The nation, society, and individuals are all concerned about these prices. For commodity housing prices, there are many factors that influence the housing prices. Apart from national regulations, factors such as lighting, layout, and environment of the houses themselves also have a certain impact on the prices, leading to significant fluctuations in the real estate market. Therefore, researching an accurate model for predicting housing prices has practical significance. It can guide residents in housing consumption and provide policy recommendations for government price regulation. Machine learning methods have become a new type of prediction method in this regard. Based on the theories of data analysis and machine learning, a dataset consisting of 2920 data points with 81 attributes was selected from the publicly available Kaggle housing dataset. The data was normalized and analyzed for feature selection. The ranking of attributes most correlated with housing prices was obtained. Subsequently, a neural network model was built, parameters were adjusted, and the trained network was used to predict housing prices. On the Kaggle leaderboard, the RMSE test result stands at 0.1198, positioning our model among the top performers among all machine learning methods.

Identifying Hard-to-Decarbonize houses from multi-source data in Cambridge, UK

As the urban population continues to expand and is expected to comprise 80% of the total population in 2050, it is crucial to ensure the sustainability and energy efficiency of cities. Among all the homes globally, Hard-to-Decarbonize (HtD) buildings are estimated to be a quarter of them. Identifying the HtD houses and proposing strategies for these houses is important to reach the global net zero target. However, the study of HtD houses has historically been neglected. Previous studies mainly focus on simulating, predicting and understanding attributes that are directly related to energy usage and efficiency. In this research, a methodology for identifying HtD buildings with publicly available data is proposed and tested in Cambridge, UK. A dataset of HtD houses in Cambridge is organized, with criteria derived from the Energy Performance Certificate (EPC), which results from detailed inspections of houses. Street view images (SVI), aeriel view images (AVI), land surface temperature (LST), and building stock data are used together for the prediction with deep learning. The classification precision for HtD buildings is able to achieve 82%. This study also explores the minimal data needed for the high-accuracy prediction of HtD houses. Results show that SVI contributes the most to the prediction.

Autocracy's long reach: explaining host country influences on transnational repression

ABSTRACT Authoritarian regimes frequently reach across borders to repress against exiled dissidents. Existing scholarship has investigated the methods and effects of transnational repression. Yet, we lack knowledge of the role that the political context of a host country and its relations to the origin country of diasporas play in incidents of transnational repression. Addressing this gap, we use a Freedom House dataset on physical acts of transnational repression (2014–2020) to study how the regime type of the host country and the regional ties between the host and origin country influence the likelihood and type of transnational repression incidents. Conducting a logistic regression analysis with yearly directed dyads, we find that to target exiles in autocratic host states perpetrators primarily rely on the cooperation of authorities, whereas in democratic host states they resort more often to direct attacks. We also show that authoritarian cooperation on transnational repression is regionally clustered: it often occurs when home and host state are situated within the same authoritarian neighbourhood, and partly also when they are members in the same regional organization. Our article reveals some of the host state conditions and relational dynamics that shape the decisions and strategies of transnational repression perpetrators.

A new algorithm for particle weighted subtraction to decrease signals from unwanted components in single particle analysis

Single particle analysis (SPA) in cryo-electron microscopy (cryo-EM) is highly used to obtain the near-atomic structure of biological macromolecules. The current methods allow users to produce high-resolution maps from many samples. However, there are still challenging cases that require extra processing to obtain high resolution. This is the case when the macromolecule of the sample is composed of different components and we want to focus just on one of them. For example, if the macromolecule is composed of several flexible subunits and we are interested in a specific one, if it is embedded in a viral capsid environment, or if it has additional components to stabilize it, such as nanodiscs. The signal from these components, which in principle we are not interested in, can be removed from the particles using a projection subtraction method. Currently, there are two projection subtraction methods used in practice and both have some limitations. In fact, after evaluating their results, we consider that the problem is still open to new solutions, as they do not fully remove the signal of the components that are not of interest. Our aim is to develop a new and more precise projection subtraction method, improving the performance of state-of-the-art methods. We tested our algorithm with data from public databases and an in–house data set. In this work, we show that the performance of our algorithm improves the results obtained by others, including the localization of small ligands, such as drugs, whose binding location is unknown a priori.

Optimization house price prediction model using gradient boosted regression trees (GBRT) and xgboost algorithm

In this rapidly advancing technological era, the demand for the real estate industry has also increased, including in the field of house price prediction. House prices fluctuate every year due to several factors such as changes in land prices, location, year of construction, infrastructure developments, and other factors. Numerous studies have been conducted on this issue. However, the challenge lies in building a proven accurate and effective model for predicting house prices with the abundance of features present in the dataset. The objective of this research is to develop a predictive model that can accurately estimate house prices based on relevant features or variables. The researcher utilizes ensemble learning techniques, combining the Gradient Boosted Regression Trees (GBRT) and XGBoost algorithms. The dataset used in this article is titled "Ames Housing dataset" obtained from Kaggle. The predictive model is then evaluated using the Root Mean Squared Error (RMSE) method. The RMSE result from a previous study that used the combination of Lasso and XGBoost was 0.11260, while the RMSE result from this research is 0.00480. This indicates a decrease in the RMSE value, indicating a lower level of error in the model. It also means that the combination of GBRT and XGBoost algorithms successfully improves the prediction accuracy of the previous research model.

Comprehensive Evaluations of Student Performance Estimation via Machine Learning

Success in student learning is the primary aim of the educational system. Artificial intelligence utilizes data and machine learning to achieve excellence in student learning. In this paper, we exploit several machine learning techniques to estimate early student performance. Two main simulations are used for the evaluation. The first simulation used the Traditional Machine Learning Classifiers (TMLCs) applied to the House dataset, and they are Gaussian Naïve Bayes (GNB), Support Vector Machine (SVM), Decision Tree (DT), Multi-Layer Perceptron (MLP), Random Forest (RF), Linear Discriminant Analysis (LDA), and Quadratic Discriminant Analysis (QDA). The best results were achieved with the MLP classifier with a division of 80% training and 20% testing, with an accuracy of 88.89%. The fusion of these seven classifiers was also applied and the highest result was equal to the MLP. Moreover, in the second simulation, the Convolutional Neural Network (CNN) was utilized and evaluated on five main datasets, namely, House, Western Ontario University (WOU), Experience Application Programming Interface (XAPI), University of California-Irvine (UCI), and Analytics Vidhya (AV). The UCI dataset was subdivided into three datasets, namely, UCI-Math, UCI-Por, and UCI-Fused. Moreover, the AV dataset has three targets which are Math, Reading, and Writing. The best accuracy results were achieved at 97.5%, 99.55%, 98.57%, 99.28%, 99.40%, 99.67%, 92.93%, 96.99%, and 96.84% for the House, WOU, XAPI, UCI-Math, UCI-Por, UCI-Fused, AV-Math, AV-Reading, and AV-Writing datasets, respectively, under the same protocol of evaluation. The system demonstrates that the proposed CNN-based method surpasses all seven conventional methods and other state-of-the-art-work.

Housing abandonment and socio-spatial inequalities: experience from a shrinking inner-city area of Incheon, South Korea

Housing abandonment is one of the most distinct characteristics of urban shrinkage. Previous research has investigated the extent and process of abandonment in the United States and Europe’s former industrialized centers. However, little is known about its features in places that have recently seen rapid urbanization. The study examines the relationship between vacant houses and a variety of urban factors, further expanding it to the perspective of spatial inequality. The multilevel logistic regression model was applied after constructing a new parcel-level dataset of vacant houses in Incheon, South Korea. The findings showed that spatially selective patterns of housing abandonment occurred at the intra-downtown level. Older, smaller, and less-accessible residential buildings were more vulnerable to abandonment. Failed large-scale redevelopment attempts fueled the emergence of vacant housing clusters. Vacant houses were associated not only with the collapse of industrial and commercial activities, but also with an unsustainable population structure. The results suggest the necessity for vacant house management plans tailored to neighborhood characteristics, especially for vacant house clusters, and a greater emphasis on socio-demographic dynamics such as age and generation in relation to urban shrinkage.

Estimate House Price Using Machine Learning

The House Price Index (HPI) is commonly used to estimate the changes in housing prices. The sale price of the house is correlated with many other factors like geographical location, size of the house, age of the house, the area and population of the neighborhood etc. There has been a considerably large number of datasets released in the literature of various locations to explore the correlation of the sale price of houses with their corresponding features. However, all the features don’t affect the sale price in equal proportion. Some features strongly correlate with each other and, while some features don’t carry any importance or are probably redundant. As a result, to explore various impacts of features on sales prices, we performed a detailed data analysis on the original house dataset. This report also comprehensively validates multiple steps of data analysis with supporting statistics and visualizations to provide an optimistic result of various features and their impact on the sales price of houses.

Optimal deep neural networks by maximization of the approximation power

We propose an optimal architecture for deep neural networks of given size. The optimal architecture obtains from maximizing the lower bound of the maximum number of linear regions approximated by a deep neural network with a ReLu activation function. The accuracy of the approximation function relies on the neural network structure characterized by the number, dependence and hierarchy between the nodes within and across layers. We show how the accuracy of the approximation improves as we optimally choose the width and depth of the network. A Monte-Carlo simulation exercise illustrates the outperformance of the optimized architecture against cross-validation methods and gridsearch for linear and nonlinear prediction models. The application of this methodology to the Boston Housing dataset confirms empirically the outperformance of our method against state-of the-art machine learning models.

Real Estate Price Prediction based on Supervised Machine Learning Scenarios

House price prediction is one of the most common supervised learning tasks in the machine learning field, which makes it a perfect criterion for the effectiveness of different learning models. From basic regression models to neural networks, countless methods have been proposed to solve the house price prediction problem. In this paper, the focus is the performance of three regression models, linear, LASSO, and ridge. There will be a selected dataset of sold houses from the open-source website. The data will be explored and visualized for a better understanding and then implement the regression models for further testing. According to the analysis, the LASSO regression model can yield the most accurate prediction with 90.15% accuracy but need a specific λ value. The linear and ridge regression yields similar predictions with close to 90% accuracy. Therefore, the most effective model for the house price prediction problem is the LASSO regression model. Overall, these results shed light on guiding further exploration of the performance of different machine learning models.