array(35) { ["id"]=> string(4) "2066" ["type"]=> string(6) "course" ["title"]=> string(88) "银行客户购买银行产品行为预测研究--基于多模型的Stacking集成学习" ["subtitle"]=> string(0) "" ["creator"]=> array(6) { ["id"]=> string(1) "3" ["nickname"]=> string(5) "admin" ["title"]=> string(1) " " ["uuid"]=> string(40) "634d3b58166bfafd4069119be97ee6bfee064c52" ["destroyed"]=> string(1) "0" ["avatar"]=> array(3) { ["small"]=> string(75) "http://www.chinadatacase.com/files/user/otherform/1711697442_1722994936.png" ["middle"]=> string(75) "http://www.chinadatacase.com/files/user/otherform/1711697442_1722994936.png" ["large"]=> string(75) "http://www.chinadatacase.com/files/user/otherform/1711697442_1722994936.png" } } ["showable"]=> string(1) "1" ["buyable"]=> string(1) "1" ["summary"]=> string(755) "

随着银行业理念的转变，从传统的“产品中心”模式过渡到“客户中心”模式，精准预测客户对银行产品的购买意愿变得至关重要。本研究采用SMOTE技术进行数据过采样，并利用Stacking集成学习方法构建预测模型，同时结合模拟退火与TPE优化算法，以实现对银行客户购买行为的二元分类预测。研究结果显示，相较于单一模型，经过Stacking集成学习优化的模型展现出更优的预测性能，其AUC值达到0.8826，召回率高达0.8160，有效识别了潜在的银行产品购买者。基于此，该模型可被应用于精准营销策略中，以增强客户忠诚度，并促进银行、合作方及客户之间的互利共赢。

" ["minPrice"]=> string(4) "0.00" ["maxPrice"]=> string(4) "0.00" ["discountId"]=> string(1) "0" ["images"]=> array(3) { ["large"]=> string(76) "http://www.chinadatacase.com/files/course/2025/03-20/1002477a232d883907.jpeg" ["middle"]=> string(76) "http://www.chinadatacase.com/files/course/2025/03-20/1002477a3ec0442636.jpeg" ["small"]=> string(76) "http://www.chinadatacase.com/files/course/2025/03-20/1002477a46f1950338.jpeg" } ["ratingNum"]=> string(1) "0" ["rating"]=> string(1) "0" ["hitNum"]=> string(3) "100" ["hotSeq"]=> string(1) "0" ["maxPriceObj"]=> array(2) { ["currency"]=> string(3) "RMB" ["amount"]=> string(4) "0.00" } ["minPriceObj"]=> array(2) { ["currency"]=> string(3) "RMB" ["amount"]=> string(4) "0.00" } ["minDisplayPrice"]=> string(4) "0.00" ["maxDisplayPrice"]=> string(4) "0.00" ["minDisplayPriceObj"]=> array(2) { ["currency"]=> string(3) "RMB" ["amount"]=> string(4) "0.00" } ["maxDisplayPriceObj"]=> array(2) { ["currency"]=> string(3) "RMB" ["amount"]=> string(4) "0.00" } ["canManage"]=> bool(false) ["peopleShowNum"]=> string(3) "119" ["isMember"]=> bool(false) ["status"]=> string(7) "created" ["orgId"]=> string(1) "1" ["orgCode"]=> string(2) "1." ["recommendWeight"]=> string(1) "0" ["recommendedTime"]=> string(1) "0" ["createdTime"]=> string(25) "2025-01-13T17:01:13+08:00" ["updatedTime"]=> string(25) "2025-12-09T16:01:35+08:00" ["product"]=> array(7) { ["id"]=> string(4) "2067" ["targetType"]=> string(6) "course" ["title"]=> string(88) "银行客户购买银行产品行为预测研究--基于多模型的Stacking集成学习" ["owner"]=> string(1) "3" ["createdTime"]=> string(10) "1736758873" ["updatedTime"]=> string(10) "1742436201" ["target"]=> array(17) { ["id"]=> string(4) "2188" ["type"]=> string(6) "normal" ["title"]=> string(88) "银行客户购买银行产品行为预测研究--基于多模型的Stacking集成学习" ["subtitle"]=> string(0) "" ["summary"]=> string(755) "

随着银行业理念的转变，从传统的“产品中心”模式过渡到“客户中心”模式，精准预测客户对银行产品的购买意愿变得至关重要。本研究采用SMOTE技术进行数据过采样，并利用Stacking集成学习方法构建预测模型，同时结合模拟退火与TPE优化算法，以实现对银行客户购买行为的二元分类预测。研究结果显示，相较于单一模型，经过Stacking集成学习优化的模型展现出更优的预测性能，其AUC值达到0.8826，召回率高达0.8160，有效识别了潜在的银行产品购买者。基于此，该模型可被应用于精准营销策略中，以增强客户忠诚度，并促进银行、合作方及客户之间的互利共赢。

" ["cover"]=> array(3) { ["large"]=> string(76) "http://www.chinadatacase.com/files/course/2025/03-20/1002477a232d883907.jpeg" ["middle"]=> string(76) "http://www.chinadatacase.com/files/course/2025/03-20/1002477a3ec0442636.jpeg" ["small"]=> string(76) "http://www.chinadatacase.com/files/course/2025/03-20/1002477a46f1950338.jpeg" } ["status"]=> string(9) "published" ["studentNum"]=> string(3) "119" ["discountType"]=> string(8) "discount" ["discount"]=> string(2) "10" ["minCoursePrice"]=> string(4) "0.00" ["maxCoursePrice"]=> string(4) "0.00" ["defaultCourseId"]=> string(4) "2194" ["productId"]=> string(4) "2067" ["goodsId"]=> string(4) "2066" ["minCoursePrice2"]=> array(2) { ["currency"]=> string(3) "RMB" ["amount"]=> string(4) "0.00" } ["maxCoursePrice2"]=> array(2) { ["currency"]=> string(3) "RMB" ["amount"]=> string(4) "0.00" } } } ["extensions"]=> array(3) { [0]=> string(8) "teachers" [1]=> string(14) "recommendGoods" [2]=> string(10) "isFavorite" } ["specs"]=> array(1) { [0]=> array(26) { ["id"]=> string(4) "2093" ["goodsId"]=> string(4) "2066" ["targetId"]=> string(4) "2194" ["title"]=> string(0) "" ["seq"]=> string(1) "1" ["status"]=> string(9) "published" ["price"]=> string(4) "0.00" ["coinPrice"]=> string(4) "0.00" ["usageMode"]=> string(7) "forever" ["usageDays"]=> string(1) "0" ["usageStartTime"]=> string(1) "0" ["usageEndTime"]=> string(1) "0" ["buyableStartTime"]=> string(1) "0" ["buyableEndTime"]=> string(1) "0" ["buyableMode"]=> NULL ["buyable"]=> string(1) "1" ["maxJoinNum"]=> string(1) "0" ["services"]=> array(0) { } ["priceObj"]=> array(2) { ["currency"]=> string(3) "RMB" ["amount"]=> string(4) "0.00" } ["displayPrice"]=> string(4) "0.00" ["displayPriceObj"]=> array(2) { ["currency"]=> string(3) "RMB" ["amount"]=> string(4) "0.00" } ["isMember"]=> bool(false) ["access"]=> array(2) { ["code"]=> string(14) "user.not_login" ["msg"]=> string(15) "用户未登录" } ["hasCertificate"]=> bool(false) ["learnUrl"]=> string(43) "http://www.chinadatacase.com/my/course/2194" ["teachers"]=> array(1) { [0]=> array(6) { ["id"]=> string(1) "3" ["nickname"]=> string(5) "admin" ["title"]=> string(1) " " ["uuid"]=> string(40) "634d3b58166bfafd4069119be97ee6bfee064c52" ["destroyed"]=> string(1) "0" ["avatar"]=> array(3) { ["small"]=> string(75) "http://www.chinadatacase.com/files/user/otherform/1711697442_1722994936.png" ["middle"]=> string(75) "http://www.chinadatacase.com/files/user/otherform/1711697442_1722994936.png" ["large"]=> string(75) "http://www.chinadatacase.com/files/user/otherform/1711697442_1722994936.png" } } } } } } 银行客户购买银行产品行为预测研究--基于多模型的Stacking集成学习 - 中国经管实验教学案例平台

中国经管实验教学案例平台

www.chinadatacase.com