Location-Centered House Price Prediction: A Multi-Task Learning Approach

Accurate house prediction is of great significance to various real estate stakeholders such as house owners, buyers, and investors. We propose a location-centered prediction framework that differs from existing work in terms of data profiling and prediction model. Regarding data profiling, we make an important observation as follows – besides the in-house features such as floor area, the location plays a critical role in house price prediction. Unfortunately, existing work either overlooked it or had a coarse grained measurement of locations. Thereby, we define and capture a fine-grained location profile powered by a diverse range of location data sources, including transportation profile, education profile, suburb profile based on census data, and facility profile. Regarding the choice of prediction model, we observe that a variety of approaches either consider the entire data for modeling, or split the entire house data and model each partition independently. However, such modeling ignores the relatedness among partitions, and for all prediction scenarios, there may not be sufficient training samples per partition for the latter approach. We address this problem by conducting a careful study of exploiting the Multi-Task Learning (MTL) model. Specifically, we map the strategies for splitting the entire house data to the ways the tasks are defined in MTL, and select specific MTL-based methods with different regularization terms to capture and exploit the relatedness among tasks. Based on real-world house transaction data collected in Melbourne, Australia, we design extensive experimental evaluations, and the results indicate a significant superiority of MTL-based methods over state-of-the-art approaches. Meanwhile, we conduct an in-depth analysis on the impact of task definitions and method selections in MTL on the prediction performance, and demonstrate that the impact of task definitions on prediction performance far exceeds that of method selections.

Particle Swarm Optimization in Machine Learning Prediction of Airbnb Hospitality Price Prediction

Particle Swarm Optimization is a metaheuristics algorithm widely used for optimization problems. This paper presents the research design and implementation of using Particle Swarm Optimization to automate the features selections in the machine learning models for Airbnb price prediction. Today, Airbnb is changing the business models of the hospitality industry globally. While a bigger impact has been given by the Airbnb community to the local economic development of each country, there has been very little effort that investigates on Airbnb pricing issue with machine learning techniques. Focusing on Airbnb Singapore, the main problem on the dataset is the low correlation of the independent variables to the hospitality price. Choosing the best combination of the independent variables is essential, which can be achieved through features selection optimization. Particle Swarm Optimization is useful to optimize the best variables combination for automating the features selection in machine learning models. By comparing the magnitude of change of the R squared values before and after the use of PSO feature selection, the result showed that the automated features selection has improved the results of all the machine learning algorithms mainly in the linear-based machine learning (Linear Regression, Lasso, Ridge). Keywords—Machine Learning, Automated Features Selection, Particle Swarm Optimization, Airbnb