Total Home Value - Automated Valuation Model

A nationwide, automated property valuation solution using a state-of-the-art, single model methodology that leverages artificial intelligence and machine learning capabilities built on Cloud technology is available with Total Home Value X from CoreLogic. A property database of more than 5.5 billion records, updated daily, captures 99.9% of U.S. properties and spans more than 50 years to produce automated valuations that can be used anywhere a current property value is relevant. While Total Home Value X uses a single model methodology, CoreLogic currently offers four different model tunings for bulk data deliveries:

- Total Home Value X for Marketing: Highest hit-rate AVM for marketing uses, such as segmentation, and validating consumer-reported home value estimates at time of application.

- Total Home Value X for Risk Management: High level of accuracy and hit rate combination for quality control, workflow, new portfolio evaluation, and other risk management uses.

- Total Home Value X for Originations: Highest level of accuracy when performing property valuation during the loan underwriting process for purchases, refinances, and home equity/HELOCs.

- Total Home Value X for Consumers: Highest hit-rate AVM with rounded values for displaying AVM information on a website for consumers.

Total Home Value X supports diverse use cases and markets, spanning mortgage, real estate, FinTech, ReTech, Government, insurance and even consumers. Total Home Value X may be ordered as a standalone bulk product delivery. Total Home Value X data is also available in the Voluntary Lien Status products from CoreLogic.

Table: THV_MODELX_V1_VIEW

Key Fields:
- CLIP
- FIPS_CODE
- SITUS_STREET_ADDRESS
- SITUS_CITY
- SITUS_STATE
- SITUS_ZIP_CODE
- LAST_MARKET_SALE_DERIVED_DATE
- LAST_MARKET_SALE_AMOUNT
- ESTIMATED_VALUE_MKTG
- CONFIDENCE_SCORE_MKTG
- FORECAST_STANDARD_DEVIATION_MKTG

搜集汇总

背景与挑战

背景概述

Total Home Value X是CoreLogic提供的全国性自动化房产估值解决方案，利用人工智能和机器学习技术，基于超过55亿条每日更新的记录，覆盖99.9%的美国房产，并支持超过50年的数据。该数据集提供四种模型调优，分别适用于营销、风险管理、贷款发放和消费者等不同场景，关键字段包括房产标识、地址、销售日期与金额以及估值相关指标。