ML之回归预测:利用13种机器学习算法对Boston(波士顿房价)数据集【13+1,506】进行回归预测(房价预测)来比较各模型性能

简介: ML之回归预测:利用13种机器学习算法对Boston(波士顿房价)数据集【13+1,506】进行回归预测(房价预测)来比较各模型性能

输出结

image.png

image.png

image.png

数据的初步查验:输出回归目标值的差异

The max target value is 50.0

The min target value is 5.0

The average target value is 22.532806324110677

LiR:The value of default measurement of LiR is 0.6763403830998702

LiR:R-squared value of DecisionTreeRegressor: 0.6763403830998702

LiR:The mean squared error of DecisionTreeRegressor: 25.096985692067726

LiR:The mean absoluate error of DecisionTreeRegressor: 3.5261239963985433

kNNR_uni:The value of default measurement of kNNR_uni is 0.6903454564606561

kNNR_uni:R-squared value of DecisionTreeRegressor: 0.6903454564606561

kNNR_uni:The mean squared error of DecisionTreeRegressor: 24.01101417322835

kNNR_uni:The mean absoluate error of DecisionTreeRegressor: 2.9680314960629928

kNNR_dis:The value of default measurement of kNNR_dis is 0.7197589970156353

kNNR_dis:R-squared value of DecisionTreeRegressor: 0.7197589970156353

kNNR_dis:The mean squared error of DecisionTreeRegressor: 21.730250160926044

kNNR_dis:The mean absoluate error of DecisionTreeRegressor: 2.8050568785108005

linear_SVR:The value of default measurement of linear_SVR is 0.651717097429608

linear_SVR:R-squared value of DecisionTreeRegressor: 0.651717097429608

linear_SVR:The mean squared error of DecisionTreeRegressor: 27.0063071393243

linear_SVR:The mean absoluate error of DecisionTreeRegressor: 3.426672916872753

poly_SVR:The value of default measurement of poly_SVR is 0.40445405800289286

poly_SVR:R-squared value of DecisionTreeRegressor: 0.4044540580028929

poly_SVR:The mean squared error of DecisionTreeRegressor: 46.1794033139523

poly_SVR:The mean absoluate error of DecisionTreeRegressor: 3.75205926674149

rbf_SVR:The value of default measurement of rbf_SVR is 0.7564068912273935

rbf_SVR:R-squared value of DecisionTreeRegressor: 0.7564068912273935

rbf_SVR:The mean squared error of DecisionTreeRegressor: 18.888525000753493

rbf_SVR:The mean absoluate error of DecisionTreeRegressor: 2.6075632979823276

DTR:The value of default measurement of DTR is 0.699313885811367

DTR:R-squared value of DecisionTreeRegressor: 0.699313885811367

DTR:The mean squared error of DecisionTreeRegressor: 23.31559055118111

DTR:The mean absoluate error of DecisionTreeRegressor: 3.1716535433070865

RFR:The value of default measurement of RFR is 0.8320900865862684

RFR:R-squared value of DecisionTreeRegressor: 0.8320900865862684

RFR:The mean squared error of DecisionTreeRegressor: 13.019952055992995

RFR:The mean absoluate error of DecisionTreeRegressor: 2.3392650918635174

ETR:The value of default measurement of ETR is 0.7595247600325825

ETR:R-squared value of DecisionTreeRegressor: 0.7595247600325824

ETR:The mean squared error of DecisionTreeRegressor: 18.646761417322832

ETR:The mean absoluate error of DecisionTreeRegressor: 2.5487401574803146

SGDR:The value of default measurement of SGDR is 0.6525677025033261

SGDR:R-squared value of DecisionTreeRegressor: 0.6525677025033261

SGDR:The mean squared error of DecisionTreeRegressor: 26.940350120746693

SGDR:The mean absoluate error of DecisionTreeRegressor: 3.524049659554681

GBR:The value of default measurement of GBR is 0.8442966156976921

GBR:R-squared value of DecisionTreeRegressor: 0.8442966156976921

GBR:The mean squared error of DecisionTreeRegressor: 12.07344198657727

GBR:The mean absoluate error of DecisionTreeRegressor: 2.2692783233003326

[LightGBM] [Warning] feature_fraction is set=0.6, colsample_bytree=1.0 will be ignored. Current value: feature_fraction=0.6

[LightGBM] [Warning] min_data_in_leaf is set=18, min_child_samples=20 will be ignored. Current value: min_data_in_leaf=18

[LightGBM] [Warning] min_sum_hessian_in_leaf is set=0.001, min_child_weight=0.001 will be ignored. Current value: min_sum_hessian_in_leaf=0.001

[LightGBM] [Warning] bagging_fraction is set=0.7, subsample=1.0 will be ignored. Current value: bagging_fraction=0.7

LGBR:The value of default measurement of LGBR is 0.824979251097139

LGBR:R-squared value of DecisionTreeRegressor: 0.824979251097139

LGBR:The mean squared error of DecisionTreeRegressor: 13.5713354452417

LGBR:The mean absoluate error of DecisionTreeRegressor: 2.3653297699911455

[0.6763403830998702, 0.6903454564606561, 0.7197589970156353, 0.651717097429608, 0.40445405800289286, 0.7564068912273935, 0.699313885811367, 0.8320900865862684, 0.7595247600325825, 0.6525677025033261, 0.8442966156976921, 0.824979251097139]

{'learning_rate': 0.09, 'max_depth': 4, 'n_estimators': 200}

rmse: 0.37116076328428194

XGBR_grid:The value of default measurement of XGBR_grid is -0.1355992935386311

XGBR_grid:R-squared value of DecisionTreeRegressor: 0.8494067182200448

XGBR_grid:The mean squared error of DecisionTreeRegressor: 11.67719810423491

XGBR_grid:The mean absoluate error of DecisionTreeRegressor: 2.156086404304805


 

设计思

image.png

image.png

 


目录
打赏
0
0
0
0
1044
分享
相关文章
【解决方案】DistilQwen2.5-R1蒸馏小模型在PAI-ModelGallery的训练、评测、压缩及部署实践
阿里云的人工智能平台 PAI,作为一站式的机器学习和深度学习平台,对DistilQwen2.5-R1模型系列提供了全面的技术支持。无论是开发者还是企业客户,都可以通过 PAI-ModelGallery 轻松实现 Qwen2.5 系列模型的训练、评测、压缩和快速部署。本文详细介绍在 PAI 平台使用 DistilQwen2.5-R1 蒸馏模型的全链路最佳实践。
基于PAI+专属网关+私网连接:构建全链路Deepseek云上私有化部署与模型调用架构
本文介绍了阿里云通过PAI+专属网关+私网连接方案,帮助企业实现DeepSeek-R1模型的私有化部署。方案解决了算力成本高、资源紧张、部署复杂和数据安全等问题,支持全链路零公网暴露及全球低延迟算力网络,最终实现技术可控、成本优化与安全可靠的AI部署路径,满足企业全球化业务需求。
PAI-Model Gallery云上一键部署阶跃星辰新模型Step1X-Edit
4月27日,阶跃星辰正式发布并开源图像编辑大模型 Step1X-Edit,性能达到开源 SOTA。Step1X-Edit模型总参数量为19B,实现 MLLM 与 DiT 的深度融合,在编辑精度与图像保真度上实现大幅提升,具备语义精准解析、身份一致性保持、高精度区域级控制三项关键能力;支持文字替换、风格迁移等11 类高频图像编辑任务类型。在最新发布的图像编辑基准 GEdit-Bench 中,Step1X-Edit 在语义一致性、图像质量与综合得分三项指标上全面领先现有开源模型,比肩 GPT-4o 与 Gemin。PAI-ModelGallery 支持Step1X-Edit一键部署方案。
【重磅发布】AllData数据中台核心功能:机器学习算法平台
杭州奥零数据科技有限公司成立于2023年,专注于数据中台业务,维护开源项目AllData并提供商业版解决方案。AllData提供数据集成、存储、开发、治理及BI展示等一站式服务,支持AI大模型应用,助力企业高效利用数据价值。
Qwen3 全尺寸模型支持通过阿里云PAI-ModelGallery 一键部署
Qwen3 是 Qwen 系列最新一代的大语言模型,提供了一系列密集(Dense)和混合专家(MOE)模型。目前,PAI 已经支持 Qwen3 全系列模型一键部署,用户可以通过 PAI-Model Gallery 快速开箱!
AI训练师入行指南(三):机器学习算法和模型架构选择
从淘金到雕琢,将原始数据炼成智能珠宝!本文带您走进数字珠宝工坊,用算法工具打磨数据金砂。从基础的经典算法到精密的深度学习模型,结合电商、医疗、金融等场景实战,手把手教您选择合适工具,打造价值连城的智能应用。掌握AutoML改装套件与模型蒸馏术,让复杂问题迎刃而解。握紧算法刻刀,为数字世界雕刻文明!
85 6
利用机器学习算法改善电商推荐系统的效率
电商行业日益竞争激烈,提升用户体验成为关键。本文将探讨如何利用机器学习算法优化电商推荐系统,通过分析用户行为数据和商品信息,实现个性化推荐,从而提高推荐效率和准确性。
327 14
实现机器学习算法时,特征选择是非常重要的一步,你有哪些推荐的方法?
实现机器学习算法时,特征选择是非常重要的一步,你有哪些推荐的方法?
223 1
解码癌症预测的密码:可解释性机器学习算法SHAP揭示XGBoost模型的预测机制
解码癌症预测的密码:可解释性机器学习算法SHAP揭示XGBoost模型的预测机制
509 0

热门文章

最新文章

AI助理

你好,我是AI助理

可以解答问题、推荐解决方案等