t22000t/house-prices-tabular
收藏资源简介:
这是一个经过整理的、可直接使用的Kaggle竞赛House Prices: Advanced Regression Techniques训练集副本(1,460行×81列),附带来自8种架构的表格建模流程的基线指标,以便新用户有参考点进行比较。数据与Kaggle的train.csv相同,源自OpenML id 42165(规范镜像)。数据集包含80个特征和1个目标变量(SalePrice,以美元计),目标变量范围为34,900美元至755,000美元(右偏,呈伽马分布),平均值为180,921美元,中位数为163,000美元。特征包括38个连续特征(35个整数和3个浮点数)和43个分类特征(对象类型),缺失值主要集中在可选特征(如PoolQC、MiscFeature、Alley等)。数据集仅包含训练数据,适用于表格回归任务,涉及房地产领域,无个人敏感信息。
This is a curated, ready-to-use copy of the training set for the Kaggle competition House Prices: Advanced Regression Techniques, with 1,460 rows × 81 columns, accompanied by baseline metrics from tabular modeling workflows across 8 distinct architectures to provide new users with a reliable reference for model performance comparison. The data is identical to Kaggle's train.csv, sourced from the canonical mirror of OpenML ID 42165. The dataset contains 80 features and 1 target variable: SalePrice, denominated in USD, which ranges from $34,900 to $755,000, is right-skewed and follows a gamma distribution, with a mean of $180,921 and a median of $163,000. The features include 38 continuous features (35 integer-type and 3 floating-point-type) and 43 categorical features (object-type), with missing values predominantly concentrated in optional features such as PoolQC, MiscFeature, Alley, and others. This dataset only contains training data, is suitable for tabular regression tasks in the real estate domain, and contains no personal sensitive information.




