地球资源数据云——数据资源详情

人力资本项目

发布时间：2026-03-17 15:32:20资源ID：2033808619429335041资源类型：免费

该数据集《Human Capital Project》主要用于监督学习任务，数据形态以表格为主，应用场景偏向医疗健康。 题目说明：Human Capital Metrics Across Health, Education, and Skills (2000 - 2025) 任务类型：表格监督学习。 建议流程：先做缺失值/异常值处理与特征编码，再比较逻辑回归、随机森林、XGBoost。 评估建议：使用分层切分或交叉验证，优先关注 F1、Recall、AUC 等分类指标。 可用文件：human_capital_project.csv。 This dataset provides a comprehensive, country - level longitudinal view of human capital indicators collected as part of the World Bank’s Human Capital Project (HCP). The HCP aims to accelerate investments in people, focusing on education, health, and skills development, to foster equitable economic growth worldwide. The data is originally sourced from the World Bank HCP datasets and has been converted from wide format to long format for easier analysis. Each row represents a unique combination of country, indicator, year, and month (for monthly observations), with the corresponding value. Data cleaning / preprocessing done: Removed columns not essential for analysis (UNIT_MEASURE, UNIT_MEASURE_LABEL, AGG_METHOD, AGG_METHOD_LABEL, OBS_STATUS, OBS_STATUS_LABEL, OBS_CONF, OBS_CONF_LABEL) Converted monthly wide columns into Year and Month columns for long format

人力资本项目

摘要概览

该数据集《Human Capital Project》主要用于监督学习任务，数据形态以表格为主，应用场景偏向医疗健康。 题目说明：Human Capital Metrics Across Health, Education, and Skills (2000 - 2025)

任务类型：表格监督学习。

建议流程：先做缺失值/异常值处理与特征编码，再比较逻辑回归、随机森林、XGBoost。

评估建议：使用分层切分或交叉验证，优先关注 F1、Recall、AUC 等分类指标。

可用文件：human_capital_project.csv。

This dataset provides a comprehensive, country - level longitudinal view of human capital indicators collected as part of the World Bank’s Human Capital Project (HCP). The HCP aims to accelerate investments in people, focusing on education, health, and skills development, to foster equitable economic growth worldwide.

The data is originally sourced from the World Bank HCP datasets and has been converted from wide format to long format for easier analysis. Each row represents a unique combination of country, indicator, year, and month (for monthly observations), with the corresponding value.

Data cleaning / preprocessing done:

Removed columns not essential for analysis (UNIT_MEASURE, UNIT_MEASURE_LABEL, AGG_METHOD, AGG_METHOD_LABEL, OBS_STATUS, OBS_STATUS_LABEL, OBS_CONF, OBS_CONF_LABEL)

Converted monthly wide columns into Year and Month columns for long format