Propensity Scores

Propensity Scores deliver nationwide property propensity score data with 6-month predictive analytics.

Using the power of machine learning techniques, models are tailored for homeowner and renter segments with over 800 predictive variables to yield accurate results and have proven to be highly predictive through rigorous tests in the market.

Propensity Score models use CoreLogic data sets, including public record data, sales transactions, local market trends, valuation and rental data.

Product Attributes:
- Likelihood of any of the following occurrences:
- Propensity of the household to apply for a new purchase loan (in the next 6 months)
- Propensity of the property to be listed for sale (in the next 6 months)
- Propensity of the property to be listed for rent (in the next 6 months)
- Propensity that an equity line will open for the property (in the next six months)
- Propensity that a property will be the subject of a refinance loan (closed, in the next six months)

Table: PROPENSITY_SCORE_V1_VIEW

Key Fields:
- CLIP
- FIPS_CODE
- SITUS_STREET_ADDRESS
- SITUS_CITY
- SITUS_STATE
- SITUS_ZIP_CODE
- NUMBER_OF_BUILDINGS
- OWNER_1_FULL_NAME
- OWNER_2_FULL_NAME
- LAST_MARKET_SALE_AMOUNT
- LIST_FOR_SALE_MODEL_PROPENSITY_SCORE
- LIST_FOR_SALE_MODEL_PROPENSITY_SCORE_RUN_DATE
- LIST_TO_RENT_MODEL_PROPENSITY_SCORE
- LIST_TO_RENT_MODEL_PROPENSITY_SCORE_RUN_DATE

搜集汇总

背景与挑战

背景概述

该数据集利用机器学习技术，基于800多个预测变量生成全国房产倾向性评分，提供6个月内房产交易、租赁、贷款等行为的预测概率。数据来源于CoreLogic的公共记录、销售交易和市场趋势等多维信息，包含关键字段如房产地址、所有者信息和各类倾向性评分。