Zillow Group has been granted a patent for a facility that uses machine learning models to value homes in a specific geographic area. The facility retrieves home sales data and creates classification trees to determine the value of a home based on its attributes and recent sales data in the area. The obtained valuation is then reported. GlobalData’s report on Zillow Group gives a 360-degree view of the company including its patenting strategy. Buy the report here.
Smarter leaders trust GlobalData
According to GlobalData’s company profile on Zillow Group, 3d modelling and rendering was a key innovation area identified from patents. Zillow Group's grant share as of September 2023 was 58%. Grant share is based on the ratio of number of grants to total number of patents.
Valuing homes in a geographic area using machine learning
A recently granted patent (Publication Number: US11769181B2) describes a computer-implemented method for generating machine learning models to value homes in a specific geographic area. The method involves retrieving home sales data for the area and creating machine learning classification trees. Each tree represents a range of values for selected home attributes and determines the information gain resulting from possible splits in these attribute ranges. The method also calculates the mean selling price for homes represented by leaf nodes in the trees.
The patent also includes additional steps to further refine the valuation process. It describes a method for scoring the classification trees based on the differences between the mean selling prices and the actual selling prices of excluded entries. This scoring helps evaluate the accuracy of the models. The patent also introduces a method for receiving attribute values for a specific home and identifying leaf nodes in the classification trees that correspond to these attribute values. The mean selling prices associated with these leaf nodes are then weighted by the tree scores and averaged to obtain a valuation for the home.
Furthermore, the patent discusses techniques for handling missing attribute values. If a particular attribute value is unavailable for a home, the method imputes a value by choosing the median value of that attribute from a set of identified homes sold in the same geographic area. The patent also introduces a blending technique to incorporate earlier-reported valuations into the obtained valuation. This blending is achieved by generating a weighted average of the obtained valuation and the earlier-reported valuation, with the earlier-reported valuation being more heavily weighted.
The patent also includes provisions for generating valuations over time and visualizing changes in home values. It describes a method for producing valuations at different time intervals and determining the extent and direction of change between these valuations. These changes can be graphically represented in a topological representation of the geographic area.
Overall, this patent presents a computer-implemented method for generating machine learning models to value homes in a specific geographic area. The method incorporates various techniques such as classification trees, scoring, imputation of missing values, blending of valuations, and visualization of changes over time. These techniques aim to improve the accuracy and reliability of home valuations in the real estate market.
To know more about GlobalData’s detailed insights on Zillow Group, buy the report here.
Data Insights
From
The gold standard of business intelligence.
Blending expert knowledge with cutting-edge technology, GlobalData’s unrivalled proprietary data will enable you to decode what’s happening in your market. You can make better informed decisions and gain a future-proof advantage over your competitors.
GlobalData, the leading provider of industry intelligence, provided the underlying data, research, and analysis used to produce this article.
GlobalData Patent Analytics tracks bibliographic data, legal events data, point in time patent ownerships, and backward and forward citations from global patenting offices. Textual analysis and official patent classifications are used to group patents into key thematic areas and link them to specific companies across the world’s largest industries.
