Real Estate Index
The price of a land interacts with and is affected dramatically by many factors such as economic, urban, and policy factors. A government entity has developed a new regulation imposing fees on undeveloped lands inside cities, with the goal of encouraging landowners to develop those empty land parcels. The new regulatory act has determined an annual fee based on the land’s “fair price.” However, because land prices vary and are difficult to interpret, the challenge is to develop a non-conventional methodology that determines the fair price for land. The method must consider all the previously mentioned factors while accounting for other regulations recently enacted in the country, which increases the stability and reliability of data. Additionally, it should handle all the issues that appear in the data, including missing information, unrelated information, and discrepancies between different data sources.
What we did
Data-driven decision-making is one of the key approaches that decision-makers try to use. Our process includes a rigorous cleaning of the available data, using various machine learning and statistical techniques. This includes separating transactions in the data that were incorrectly labeled by the data owner. In addition, other statistical techniques are used to eliminate outliers, remove inflation, and discard unreal transactions. These different cleaning steps improve the quality of the data, which is necessary to get reliable results. Further, the cleaned real estate transactions are processed to produce a representative set of indicators in various geographical resolutions. The indicators are extracted using different statistical and clustering techniques that process the various available data sources. Furthermore, Intelmatix decision-making tools empower the decision-maker to analyze the data easily.
The valuation model analysis introduced three main concepts: the neighborhood index, the zoning index, and the city index. The neighborhood index gives an indication of the neighborhood’s ranking within the city by showing the distribution of prices for the different neighborhoods. Being the most granular level, it represented one of the main indicators that have been used to define the fair price of a specific property within a neighborhood. In many cases, agencies want to avoid being market makers. Therefore, a zoning index has been developed to group the city into zones based on the full profiles of neighborhoods (neighborhoods with similar price ranges). Finally, the city index provides an indication of the most common price across the city, which allows regional comparison across different cities. Furthermore, all components have been integrated into an interactive visualization tool to help the decision-makers evaluate land parcels across various areas in the city.