COD风控模型
LIam
2019/1/14
COD风控模型
数据范围
特征纬度包括两个方面：1.订单信息 2. 用户信息
- 物流状态数据：logistics_workflow_union
- 所以订单数据：sale_order_union
- 用户信息：wp_cid_collect_union
- 衍生字段：用户过去的购买情况，比如过去的签收率
时间范围
- 用户信息的构造范围是2018年10月1�号之前
- 使用2018年10月1号到2018年10月31号的订单数据作为训练
数据
字段含义如下： - 订单名字 - 订单创建时间 - 订单更新时间 - 物流状态 - 用户的id - 用户的州 - 下单总额 - 客户端版本 - 客户端型号 - 手机设备 - 性别 - 下过多少单 - 成功支付过多少订单 - 成功支付过多少COD订单 - 签收过多少COD订单 - COD 订单的签收率 - 下单的总金额 - COD成功支付订单的总金额 - COD签收订单的总金额
一共35w条数据，坏样本比例大约为 24%
数据分析
下面是挑出一些比较有区分能力的特征的分析结果：
- 不同的州
| count | count_distr | good | bad | badprob | woe | bin_iv | total_iv |
|---|---|---|---|---|---|---|---|
| 56558 | 0.1563767 | 47417 | 9141 | 0.1616217 | -0.4893475 | 0.0326334 | 0.1031834 |
| 74501 | 0.2059871 | 59091 | 15410 | 0.2068429 | -0.1871984 | 0.0068634 | 0.1031834 |
| 133546 | 0.3692400 | 102691 | 30855 | 0.2310440 | -0.0455621 | 0.0007574 | 0.1031834 |
| 75636 | 0.2091252 | 53330 | 22306 | 0.2949125 | 0.2852202 | 0.0182460 | 0.1031834 |
| 21437 | 0.0592710 | 12622 | 8815 | 0.4112049 | 0.7978770 | 0.0446833 | 0.1031834 |
坏样本比较低的州包括：
## [1] "punjab,up,telangana,ponducherry,jalandhar,maharastra,haryana,firozpur,Puducherry,Daman and Diu,Goa,Meghalaya,Manipur,Mizoram,Chandigarh,Kerala,Delhi"## [1] "一共 17 个州"坏样本比较高的州包括：
## [1] "Sikkim,Assam,Tripura,rajsthan,dadra and nagar haveli"## [1] "一共 5 个州"建立模型并评估
##
## Call:
## roc.default(response = tmp[, 2], predictor = tmp[, 1], plot = TRUE, print.thres = TRUE, print.auc = TRUE)
##
## Data: tmp[, 1] in 55030 controls (tmp[, 2] 0) < 17305 cases (tmp[, 2] 1).
## Area under the curve: 0.6645直接构建决策树,获取规则
##
## Call:
## roc.default(response = (testtree %>% filter(gender != "U"))$label, predictor = treepre[, 2], plot = TRUE, print.thres = TRUE, print.auc = TRUE)
##
## Data: treepre[, 2] in 55030 controls ((testtree %>% filter(gender != "U"))$label 0) < 17305 cases ((testtree %>% filter(gender != "U"))$label 1).
## Area under the curve: 0.6149##
## Call:
## C5.0.formula(formula = label ~ ., data = traintree[, -5], rules =
## T, control = tc)
##
##
## C5.0 [Release 2.07 GPL Edition] Tue Jan 15 14:31:26 2019
## -------------------------------
##
## Class specified by attribute `outcome'
##
## Read 289343 cases (15 attributes) from undefined.data
##
## Rules:
##
## Rule 1: (1508/127, lift 1.2)
## shipping_state = Goa
## gender = F
## -> class 0 [0.915]
##
## Rule 2: (1821/214, lift 1.2)
## shipping_state = Manipur
## gender = F
## -> class 0 [0.882]
##
## Rule 3: (3552/506, lift 1.1)
## client_version = 5.0.1
## order_paysuccess > 2
## -> class 0 [0.857]
##
## Rule 4: (27871/4941, lift 1.1)
## client_os = iOS
## -> class 0 [0.823]
##
## Rule 5: (62045/11085, lift 1.1)
## client_version = 5.0.3
## -> class 0 [0.821]
##
## Rule 6: (141329/25424, lift 1.1)
## gender = F
## -> class 0 [0.820]
##
## Rule 7: (70297/12667, lift 1.1)
## sig_rate > 0.4635762
## -> class 0 [0.820]
##
## Rule 8: (6615/1340, lift 1.0)
## client_version = 4.8.6
## -> class 0 [0.797]
##
## Rule 9: (31857/6517, lift 1.0)
## client_version = 4.9.2
## -> class 0 [0.795]
##
## Rule 10: (91266/19576, lift 1.0)
## amount_total <= 4.43
## -> class 0 [0.785]
##
## Rule 11: (11066/2501, lift 1.0)
## client_version = 4.8.5
## -> class 0 [0.774]
##
## Rule 12: (207727/51301, lift 1.0)
## cod_paysuccess <= 0
## -> class 0 [0.753]
##
## Rule 13: (9/1, lift 3.4)
## client_version = 4.2.2
## cod_paysuccess > 0
## sig_rate <= 0.4635762
## -> class 1 [0.818]
##
## Rule 14: (7/1, lift 3.3)
## client_version = 3.6.2
## gender = M
## cod_paysuccess > 0
## sig_rate <= 0.6832298
## -> class 1 [0.778]
##
## Rule 15: (25/6, lift 3.1)
## shipping_state = Delhi
## gender = F
## cod_paysuccess > 1
## cod_paysuccess <= 7
## sig_rate <= 0.4635762
## cod_paysuccess_amount > 35.4
## -> class 1 [0.741]
##
## Rule 16: (31/9, lift 2.9)
## shipping_state = Odisha
## gender = F
## cod_paysuccess > 1
## cod_paysuccess <= 7
## sig_rate <= 0.4635762
## -> class 1 [0.697]
##
## Rule 17: (11/3, lift 2.9)
## shipping_state = Chhattisgarh
## gender = F
## cod_paysuccess > 1
## sig_rate <= 0.4635762
## -> class 1 [0.692]
##
## Rule 18: (108/35, lift 2.8)
## gender = F
## cod_paysuccess > 7
## sig_rate <= 0.4635762
## -> class 1 [0.673]
##
## Rule 19: (168/58, lift 2.7)
## shipping_state = Karnataka
## cod_paysuccess > 1
## cod_delivered <= 0
## -> class 1 [0.653]
##
## Rule 20: (101/35, lift 2.7)
## gender = M
## cod_paysuccess > 9
## sig_rate <= 0.6832298
## -> class 1 [0.650]
##
## Rule 21: (79/28, lift 2.7)
## shipping_state = Punjab
## cod_paysuccess > 1
## cod_paysuccess <= 7
## sig_rate <= 0.4635762
## -> class 1 [0.642]
##
## Rule 22: (64/25, lift 2.5)
## shipping_state = Assam
## gender = F
## cod_paysuccess > 1
## sig_rate <= 0.4635762
## -> class 1 [0.606]
##
## Rule 23: (1121/445, lift 2.5)
## amount_total > 4
## client_version = 4.7.7
## gender = M
## cod_paysuccess > 0
## sig_rate <= 0.4635762
## -> class 1 [0.603]
##
## Rule 24: (25/10, lift 2.5)
## client_version = 4.3.3
## gender = M
## cod_paysuccess > 0
## sig_rate <= 0.4635762
## -> class 1 [0.593]
##
## Rule 25: (169/70, lift 2.4)
## client_os = android
## client_version = 5.0.1
## gender = M
## order_paysuccess <= 2
## cod_paysuccess > 0
## sig_rate <= 0.6832298
## -> class 1 [0.585]
##
## Rule 26: (599/253, lift 2.4)
## client_version = 4.6.7
## gender = M
## cod_paysuccess > 0
## sig_rate <= 0.4635762
## -> class 1 [0.577]
##
## Rule 27: (2524/1134, lift 2.3)
## cod_paysuccess > 1
## sig_rate <= 0.4635762
## -> class 1 [0.551]
##
## Rule 28: (85/39, lift 2.3)
## client_version = 4.7.6
## gender = M
## cod_paysuccess > 0
## sig_rate <= 0.4635762
## -> class 1 [0.540]
##
## Rule 29: (10396/5519, lift 2.0)
## client_os = android
## cod_paysuccess > 0
## sig_rate <= 0.4635762
## -> class 1 [0.469]
##
## Rule 30: (147702/103984, lift 1.2)
## gender = M
## -> class 1 [0.296]
##
## Default class: 0
##
##
## Evaluation on training data (289343 cases):
##
## Rules
## ----------------
## No Errors
##
## 30 68478(23.7%) <<
##
##
## (a) (b) <-classified as
## ----- -----
## 218513 1608 (a): class 0
## 66870 2352 (b): class 1
##
##
## Attribute usage:
##
## 99.89% gender
## 75.51% cod_paysuccess
## 40.49% client_version
## 31.90% amount_total
## 27.97% sig_rate
## 13.24% client_os
## 1.29% order_paysuccess
## 1.28% shipping_state
## 0.06% cod_delivered
## 0.01% cod_paysuccess_amount
##
##
## Time: 7.3 secs对老客户是否会拒收进行建模
##
## Call:
## roc.default(response = tmp[, 2], predictor = tmp[, 1], plot = TRUE, print.thres = TRUE, print.auc = TRUE)
##
## Data: tmp[, 1] in 21092 controls (tmp[, 2] 0) < 5986 cases (tmp[, 2] 1).
## Area under the curve: 0.6954构建决策树
##
## Call:
## roc.default(response = (testtree %>% filter(gender != "U"))$label, predictor = treepre[, 2], plot = TRUE, print.thres = TRUE, print.auc = TRUE)
##
## Data: treepre[, 2] in 21092 controls ((testtree %>% filter(gender != "U"))$label 0) < 5986 cases ((testtree %>% filter(gender != "U"))$label 1).
## Area under the curve: 0.6041##
## Call:
## C5.0.formula(formula = label ~ ., data = traintree[, -5], rules =
## T, control = tc)
##
##
## C5.0 [Release 2.07 GPL Edition] Tue Jan 15 14:31:49 2019
## -------------------------------
##
## Class specified by attribute `outcome'
##
## Read 108317 cases (15 attributes) from undefined.data
##
## Rules:
##
## Rule 1: (37288/7596, lift 1.0)
## cod_paysuccess > 1
## -> class 0 [0.796]
##
## Rule 2: (71029/16349, lift 1.0)
## cod_paysuccess <= 1
## -> class 0 [0.770]
##
## Rule 3: (4, lift 3.8)
## client_version = 4.0.3
## gender = F
## cod_paysuccess > 1
## sig_rate <= 0.4635762
## -> class 1 [0.833]
##
## Rule 4: (4, lift 3.8)
## client_version = 4.2.2
## gender = M
## cod_paysuccess > 0
## sig_rate <= 0.7017544
## -> class 1 [0.833]
##
## Rule 5: (8/1, lift 3.6)
## client_version = 4.1.1
## gender = M
## cod_paysuccess > 0
## sig_rate <= 0.4635762
## -> class 1 [0.800]
##
## Rule 6: (3, lift 3.6)
## client_version = 4.4.0
## cod_paysuccess > 0
## sig_rate <= 0.4635762
## -> class 1 [0.800]
##
## Rule 7: (23/5, lift 3.4)
## client_version = 4.6.5
## gender = M
## cod_paysuccess > 0
## cod_paysuccess <= 1
## sig_rate <= 0.7017544
## -> class 1 [0.760]
##
## Rule 8: (6/1, lift 3.4)
## client_version = 4.8.2
## gender = F
## cod_paysuccess > 1
## sig_rate <= 0.4635762
## -> class 1 [0.750]
##
## Rule 9: (13/3, lift 3.3)
## client_version = 4.3.4
## gender = M
## cod_paysuccess > 0
## sig_rate <= 0.7017544
## -> class 1 [0.733]
##
## Rule 10: (9/2, lift 3.3)
## client_version = 4.6.1
## cod_paysuccess > 0
## sig_rate <= 0.7017544
## -> class 1 [0.727]
##
## Rule 11: (45/12, lift 3.3)
## amount_total > 9.63
## client_version = 4.5.4
## gender = M
## cod_paysuccess > 0
## sig_rate <= 0.4635762
## -> class 1 [0.723]
##
## Rule 12: (8/2, lift 3.2)
## client_version = 3.6.2
## cod_paysuccess > 0
## sig_rate <= 0.7017544
## -> class 1 [0.700]
##
## Rule 13: (21/6, lift 3.1)
## client_version = 4.3.3
## cod_paysuccess > 1
## sig_rate <= 0.4635762
## -> class 1 [0.696]
##
## Rule 14: (4/1, lift 3.0)
## client_version = 4.7.4
## sig_rate <= 0.4635762
## -> class 1 [0.667]
##
## Rule 15: (220/77, lift 2.9)
## client_version = 4.6.7
## cod_paysuccess > 1
## sig_rate <= 0.4635762
## -> class 1 [0.649]
##
## Rule 16: (1007/356, lift 2.9)
## amount_total > 4.43
## gender = M
## cod_paysuccess > 1
## sig_rate <= 0.4635762
## -> class 1 [0.646]
##
## Rule 17: (81/29, lift 2.9)
## client_version = 4.4.5
## gender = M
## cod_paysuccess > 0
## sig_rate <= 0.4635762
## -> class 1 [0.639]
##
## Rule 18: (119/45, lift 2.8)
## client_version = 4.3.5
## gender = M
## cod_paysuccess > 0
## sig_rate <= 0.4635762
## -> class 1 [0.620]
##
## Rule 19: (471/190, lift 2.7)
## client_version = 4.7.7
## cod_paysuccess > 1
## sig_rate <= 0.4635762
## -> class 1 [0.596]
##
## Rule 20: (1126/459, lift 2.7)
## amount_total > 4
## client_version = 4.7.7
## gender = M
## cod_paysuccess > 0
## sig_rate <= 0.4635762
## -> class 1 [0.592]
##
## Rule 21: (596/251, lift 2.6)
## client_version = 4.6.7
## gender = M
## cod_paysuccess > 0
## sig_rate <= 0.4635762
## -> class 1 [0.579]
##
## Rule 22: (304/128, lift 2.6)
## client_version = 4.9.2
## cod_paysuccess > 1
## sig_rate <= 0.4635762
## -> class 1 [0.578]
##
## Rule 23: (72/31, lift 2.6)
## client_version = 4.7.6
## gender = M
## cod_paysuccess > 0
## cod_paysuccess <= 1
## sig_rate <= 0.7017544
## -> class 1 [0.568]
##
## Rule 24: (108/47, lift 2.5)
## client_os = android
## client_version = 5.0.1
## gender = M
## cod_paysuccess > 0
## cod_paysuccess <= 1
## sig_rate <= 0.7017544
## -> class 1 [0.564]
##
## Rule 25: (14/6, lift 2.5)
## client_version = missing
## gender = M
## sig_rate <= 0.7017544
## -> class 1 [0.563]
##
## Rule 26: (57/25, lift 2.5)
## client_version = 4.7.3
## gender = M
## cod_paysuccess > 0
## cod_paysuccess <= 1
## sig_rate <= 0.7017544
## -> class 1 [0.559]
##
## Rule 27: (45262/31677, lift 1.4)
## sig_rate <= 0.7017544
## -> class 1 [0.300]
##
## Default class: 0
##
##
## Evaluation on training data (108317 cases):
##
## Rules
## ----------------
## No Errors
##
## 27 23263(21.5%) <<
##
##
## (a) (b) <-classified as
## ---- ----
## 82919 1453 (a): class 0
## 21810 2135 (b): class 1
##
##
## Attribute usage:
##
## 100.00% cod_paysuccess
## 41.79% sig_rate
## 2.78% client_version
## 2.75% gender
## 1.84% amount_total
## 0.10% client_os
##
##
## Time: 2.2 secs对新客户是否会拒收进行建模
Raw_data20181001old <- sample_dataset1 %>% select(-order_name,-create_time,
-update_time,-user_id,- rn,
- urn,-uid,-state) %>% filter(is.na(order_num))
Labelold <- as.numeric(as.factor(Raw_data20181001old$status))-1
# Label
# 0 1
# 105464 29931 bad rate 0.2210643
# woeold <- IV_WOE(Raw_data = Raw_data20181001old[,-1],Label = Labelold,Woe_t = T)
load("/Users/milin/COD\ 建模/woenew.RData")
woet <- woeold[[1]] %>% select(-order_paysuccess,-cod_paysuccess,-cod_delivered,
-sig_rate,-allorder_amount,
-cod_paysuccess_amount,
-cod_delivered_amount)
woet$label <- as.factor(woet$label)
tmp <- SplitSample(Raw_data = woet[,-1],Label = woet$label,rate = 0.8)
train <- tmp[[1]]
test <- tmp[[2]]
l <- glm(label ~.,data = train, family = "binomial")
pre <- predict(l,newdata = test,type = 'response')
tmp <- data.frame(pre,test$label)
library("pROC")
roc(response = tmp[,2], predictor = tmp[,1],plot=TRUE, print.thres=TRUE, print.auc=TRUE)
##
## Call:
## roc.default(response = tmp[, 2], predictor = tmp[, 1], plot = TRUE, print.thres = TRUE, print.auc = TRUE)
##
## Data: tmp[, 1] in 33937 controls (tmp[, 2] 0) < 11319 cases (tmp[, 2] 1).
## Area under the curve: 0.6521对新客户进行构建决策树
## [1] "样本标签的比例为: 2.99821542158456"
## [1] "平衡之后的标签比例为: 1"##
## Call:
## C5.0.formula(formula = label ~ ., data = traintree[, -5], rules =
## T, control = tc)
##
## Rule-Based Model
## Number of samples: 90554
## Number of predictors: 14
##
## Number of Rules: 80
##
## Non-standard options: minimum number of cases: 20
##
## Call:
## roc.default(response = (testtree %>% filter(gender != "U"))$label, predictor = treepre[, 2], plot = TRUE, print.thres = TRUE, print.auc = TRUE)
##
## Data: treepre[, 2] in 11319 controls ((testtree %>% filter(gender != "U"))$label 0) < 11319 cases ((testtree %>% filter(gender != "U"))$label 1).
## Area under the curve: 0.6117##
## Call:
## C5.0.formula(formula = label ~ ., data = traintree[, -5], rules =
## T, control = tc)
##
##
## C5.0 [Release 2.07 GPL Edition] Tue Jan 15 14:32:15 2019
## -------------------------------
##
## Class specified by attribute `outcome'
##
## Read 90554 cases (15 attributes) from undefined.data
##
## Rules:
##
## Rule 1: (8, lift 1.8)
## client_version = 4.6.1
## -> class 0 [0.900]
##
## Rule 2: (69/8, lift 1.7)
## shipping_state = Goa
## client_version = 5.0.3
## -> class 0 [0.873]
##
## Rule 3: (4, lift 1.7)
## client_version = 4.5.1
## -> class 0 [0.833]
##
## Rule 4: (59/11, lift 1.6)
## shipping_state = Puducherry
## client_os = android
## gender = F
## -> class 0 [0.803]
##
## Rule 5: (45/10, lift 1.5)
## shipping_state = Rajasthan
## client_version = 4.6.7
## gender = F
## -> class 0 [0.766]
##
## Rule 6: (44/10, lift 1.5)
## client_version = 5.0.3
## gender = M
## ipnum > 6
## -> class 0 [0.761]
##
## Rule 7: (313/75, lift 1.5)
## shipping_state = Goa
## gender = F
## -> class 0 [0.759]
##
## Rule 8: (2, lift 1.5)
## client_version = 4.6.4
## -> class 0 [0.750]
##
## Rule 9: (2, lift 1.5)
## shipping_state = punjab
## -> class 0 [0.750]
##
## Rule 10: (26/6, lift 1.5)
## shipping_state = Andhra Pradesh
## client_os = android
## client_version = 5.0.1
## -> class 0 [0.750]
##
## Rule 11: (106/27, lift 1.5)
## shipping_state = Uttarakhand
## client_version = 5.0.3
## -> class 0 [0.741]
##
## Rule 12: (741/198, lift 1.5)
## shipping_state = Delhi
## client_version = 5.0.1
## -> class 0 [0.732]
##
## Rule 13: (143/40, lift 1.4)
## client_version = 4.9.1
## gender = F
## -> class 0 [0.717]
##
## Rule 14: (1854/535, lift 1.4)
## client_version = 5.0.1
## gender = F
## -> class 0 [0.711]
##
## Rule 15: (153/45, lift 1.4)
## shipping_state = Kerala
## client_version = 5.0.3
## -> class 0 [0.703]
##
## Rule 16: (3018/898, lift 1.4)
## shipping_state = Delhi
## gender = F
## -> class 0 [0.702]
##
## Rule 17: (437/130, lift 1.4)
## shipping_state = Meghalaya
## client_os = android
## gender = F
## -> class 0 [0.702]
##
## Rule 18: (266/79, lift 1.4)
## shipping_state = West Bengal
## amount_total <= 8.88
## client_version = 5.0.3
## ipnum <= 6
## -> class 0 [0.701]
##
## Rule 19: (61/18, lift 1.4)
## shipping_state = Karnataka
## client_os = android
## client_version = 4.8.3
## gender = F
## ipnum > 2
## -> class 0 [0.698]
##
## Rule 20: (50/15, lift 1.4)
## client_version = 5.0.0
## gender = F
## -> class 0 [0.692]
##
## Rule 21: (1755/541, lift 1.4)
## shipping_state = Delhi
## client_version = 5.0.3
## -> class 0 [0.692]
##
## Rule 22: (6182/1958, lift 1.4)
## client_version = 5.0.3
## gender = F
## -> class 0 [0.683]
##
## Rule 23: (958/312, lift 1.3)
## amount_total <= 1.97
## -> class 0 [0.674]
##
## Rule 24: (280/92, lift 1.3)
## shipping_state = Maharashtra
## client_version = 5.0.1
## ipnum > 1
## -> class 0 [0.670]
##
## Rule 25: (316/104, lift 1.3)
## shipping_state = Andhra Pradesh
## client_version = 5.0.3
## -> class 0 [0.670]
##
## Rule 26: (934/312, lift 1.3)
## shipping_state = Delhi
## client_version = 4.9.2
## -> class 0 [0.666]
##
## Rule 27: (66/22, lift 1.3)
## shipping_state = Haryana
## client_version = 4.9.2
## ipnum > 1
## -> class 0 [0.662]
##
## Rule 28: (406/139, lift 1.3)
## shipping_state = Manipur
## -> class 0 [0.657]
##
## Rule 29: (123/42, lift 1.3)
## shipping_state = Chhattisgarh
## amount_total <= 9.29
## client_version = 4.7.7
## gender = F
## -> class 0 [0.656]
##
## Rule 30: (113/39, lift 1.3)
## shipping_state = West Bengal
## client_version = 4.8.3
## gender = F
## ipnum > 1
## -> class 0 [0.652]
##
## Rule 31: (216/75, lift 1.3)
## shipping_state = Punjab
## amount_total <= 4.48
## client_version = 5.0.3
## -> class 0 [0.651]
##
## Rule 32: (4472/1572, lift 1.3)
## shipping_state = Maharashtra
## gender = F
## -> class 0 [0.648]
##
## Rule 33: (1157/409, lift 1.3)
## shipping_state = Kerala
## gender = F
## -> class 0 [0.646]
##
## Rule 34: (823/291, lift 1.3)
## client_version = 4.8.6
## gender = F
## -> class 0 [0.646]
##
## Rule 35: (705/252, lift 1.3)
## shipping_state = Tamil Nadu
## client_version = 5.0.3
## ipnum <= 3
## -> class 0 [0.642]
##
## Rule 36: (273/98, lift 1.3)
## shipping_state = Chandigarh
## client_os = android
## gender = F
## -> class 0 [0.640]
##
## Rule 37: (625/226, lift 1.3)
## shipping_state = Uttarakhand
## gender = F
## -> class 0 [0.638]
##
## Rule 38: (184/67, lift 1.3)
## shipping_state = Andhra Pradesh
## client_version = 4.9.2
## -> class 0 [0.634]
##
## Rule 39: (3528/1302, lift 1.3)
## client_version = 4.9.2
## gender = F
## -> class 0 [0.631]
##
## Rule 40: (2070/774, lift 1.3)
## shipping_state = Tamil Nadu
## gender = F
## -> class 0 [0.626]
##
## Rule 41: (288/108, lift 1.2)
## shipping_state = Mizoram
## -> class 0 [0.624]
##
## Rule 42: (551/207, lift 1.2)
## shipping_state = Tamil Nadu
## client_version = 4.9.2
## -> class 0 [0.624]
##
## Rule 43: (6616/2499, lift 1.2)
## amount_total <= 5.58
## gender = F
## ipnum > 1
## -> class 0 [0.622]
##
## Rule 44: (416/157, lift 1.2)
## shipping_state = Delhi
## client_version = 4.8.5
## -> class 0 [0.622]
##
## Rule 45: (1323/501, lift 1.2)
## client_version = 4.8.5
## gender = F
## -> class 0 [0.621]
##
## Rule 46: (10063/3840, lift 1.2)
## amount_total <= 3.96
## gender = F
## ipnum <= 4
## -> class 0 [0.618]
##
## Rule 47: (50/19, lift 1.2)
## client_version = 4.0.3
## -> class 0 [0.615]
##
## Rule 48: (1389/536, lift 1.2)
## shipping_state = Punjab
## client_os = android
## gender = F
## -> class 0 [0.614]
##
## Rule 49: (243/94, lift 1.2)
## shipping_state = Tamil Nadu
## client_version = 4.8.5
## -> class 0 [0.612]
##
## Rule 50: (2396/930, lift 1.2)
## shipping_state = Maharashtra
## client_version = 5.0.3
## ipnum <= 3
## -> class 0 [0.612]
##
## Rule 51: (121/47, lift 1.2)
## shipping_state = Rajasthan
## client_version = 5.0.3
## ipnum <= 1
## -> class 0 [0.610]
##
## Rule 52: (7125/2831, lift 1.2)
## client_os = iOS
## -> class 0 [0.603]
##
## Rule 53: (737/293, lift 1.2)
## client_version = 4.8.6
## ipnum > 1
## -> class 0 [0.602]
##
## Rule 54: (646/259, lift 1.2)
## amount_total <= 17.46
## client_version = 4.6.7
## gender = F
## ipnum <= 1
## -> class 0 [0.599]
##
## Rule 55: (179/74, lift 1.2)
## shipping_state = Punjab
## client_version = 4.8.5
## -> class 0 [0.586]
##
## Rule 56: (525/223, lift 1.1)
## shipping_state = Himachal Pradesh
## -> class 0 [0.575]
##
## Rule 57: (716/320, lift 1.1)
## shipping_state = Madhya Pradesh
## amount_total <= 11.19
## ipnum <= 1
## -> class 0 [0.553]
##
## Rule 58: (7439/3398, lift 1.1)
## shipping_state = Gujarat
## -> class 0 [0.543]
##
## Rule 59: (129/26, lift 1.6)
## shipping_state = Tripura
## client_version = 4.7.7
## -> class 1 [0.794]
##
## Rule 60: (222/52, lift 1.5)
## shipping_state = Tripura
## amount_total > 3.96
## -> class 1 [0.763]
##
## Rule 61: (2, lift 1.5)
## client_version = 3.3.1
## -> class 1 [0.750]
##
## Rule 62: (416/118, lift 1.4)
## shipping_state = Sikkim
## client_os = android
## ipnum <= 6
## -> class 1 [0.715]
##
## Rule 63: (418/120, lift 1.4)
## shipping_state = Sikkim
## client_os = android
## -> class 1 [0.712]
##
## Rule 64: (56/17, lift 1.4)
## shipping_state = Arunachal Pradesh
## client_version = 4.9.2
## -> class 1 [0.690]
##
## Rule 65: (54/17, lift 1.4)
## shipping_state = Uttar Pradesh
## client_version = 5.0.1
## gender = M
## -> class 1 [0.679]
##
## Rule 66: (482/155, lift 1.4)
## shipping_state = Bihar
## gender = M
## ipnum <= 6
## -> class 1 [0.678]
##
## Rule 67: (93/31, lift 1.3)
## client_os = missing
## -> class 1 [0.663]
##
## Rule 68: (21/7, lift 1.3)
## client_version = 4.2.2
## -> class 1 [0.652]
##
## Rule 69: (80/29, lift 1.3)
## gender = missing
## -> class 1 [0.634]
##
## Rule 70: (289/106, lift 1.3)
## shipping_state = Punjab
## amount_total > 4.48
## gender = M
## ipnum > 2
## -> class 1 [0.632]
##
## Rule 71: (20388/7519, lift 1.3)
## client_version = 4.7.7
## gender = M
## -> class 1 [0.631]
##
## Rule 72: (200/76, lift 1.2)
## client_version = 4.8.1
## gender = M
## -> class 1 [0.619]
##
## Rule 73: (8291/3238, lift 1.2)
## amount_total > 1.97
## client_os = android
## client_version = 4.8.3
## gender = M
## -> class 1 [0.609]
##
## Rule 74: (1245/490, lift 1.2)
## client_version = 4.6.7
## gender = M
## -> class 1 [0.606]
##
## Rule 75: (630/253, lift 1.2)
## shipping_state = Arunachal Pradesh
## -> class 1 [0.598]
##
## Rule 76: (157/63, lift 1.2)
## shipping_state = Nagaland
## client_version = 4.7.7
## -> class 1 [0.597]
##
## Rule 77: (861/352, lift 1.2)
## client_version = 4.8.2
## gender = M
## -> class 1 [0.591]
##
## Rule 78: (387/160, lift 1.2)
## client_version = 4.8.3
## ipnum > 4
## -> class 1 [0.586]
##
## Rule 79: (147/63, lift 1.1)
## client_version = 4.6.5
## -> class 1 [0.570]
##
## Rule 80: (83336/40952, lift 1.0)
## client_os = android
## -> class 1 [0.509]
##
## Default class: 0
##
##
## Evaluation on training data (90554 cases):
##
## Rules
## ----------------
## No Errors
##
## 80 35744(39.5%) <<
##
##
## (a) (b) <-classified as
## ---- ----
## 25242 20035 (a): class 0
## 15709 29568 (b): class 1
##
##
## Attribute usage:
##
## 100.00% client_os
## 65.10% gender
## 58.20% client_version
## 35.53% shipping_state
## 25.78% amount_total
## 21.21% ipnum
##
##
## Time: 2.2 secs预测用户是否会取消
##
## Call:
## roc.default(response = tmp[, 2], predictor = tmp[, 1], plot = TRUE, print.thres = TRUE, print.auc = TRUE)
##
## Data: tmp[, 1] in 1823 controls (tmp[, 2] 0) < 685 cases (tmp[, 2] 1).
## Area under the curve: 0.6744构建决策树，尝试获得规则
## [1] "样本标签的比例为: 2.66005252407353"
## [1] "平衡之后的标签比例为: 1"##
## Call:
## C5.0.formula(formula = label ~ ., data = traintree[, -5], rules =
## T, control = tc)
##
## Rule-Based Model
## Number of samples: 5484
## Number of predictors: 14
##
## Number of Rules: 9
##
## Non-standard options: minimum number of cases: 20
##
## Call:
## roc.default(response = (testtree %>% filter(gender != "U"))$label, predictor = treepre[, 2], plot = TRUE, print.thres = TRUE, print.auc = TRUE)
##
## Data: treepre[, 2] in 685 controls ((testtree %>% filter(gender != "U"))$label 0) < 685 cases ((testtree %>% filter(gender != "U"))$label 1).
## Area under the curve: 0.5522##
## Call:
## C5.0.formula(formula = label ~ ., data = traintree[, -5], rules =
## T, control = tc)
##
##
## C5.0 [Release 2.07 GPL Edition] Tue Jan 15 14:32:20 2019
## -------------------------------
##
## Class specified by attribute `outcome'
##
## Read 5484 cases (15 attributes) from undefined.data
##
## Rules:
##
## Rule 1: (28/5, lift 1.6)
## shipping_state = Goa
## -> class 0 [0.800]
##
## Rule 2: (1189/430, lift 1.3)
## allorder_amount > 5.29
## -> class 0 [0.638]
##
## Rule 3: (77/29, lift 1.2)
## shipping_state = Delhi
## client_version = 4.8.3
## -> class 0 [0.620]
##
## Rule 4: (273/112, lift 1.2)
## client_version = 5.0.1
## -> class 0 [0.589]
##
## Rule 5: (1038/441, lift 1.1)
## client_version = 5.0.3
## -> class 0 [0.575]
##
## Rule 6: (606/273, lift 1.1)
## client_version = 4.9.2
## -> class 0 [0.549]
##
## Rule 7: (283/105, lift 1.3)
## shipping_state = Telangana
## allorder_amount <= 5.29
## -> class 1 [0.628]
##
## Rule 8: (390/161, lift 1.2)
## shipping_state = Gujarat
## allorder_amount <= 5.29
## -> class 1 [0.587]
##
## Rule 9: (4295/1983, lift 1.1)
## allorder_amount <= 5.29
## -> class 1 [0.538]
##
## Default class: 0
##
##
## Evaluation on training data (5484 cases):
##
## Rules
## ----------------
## No Errors
##
## 9 2284(41.6%) <<
##
##
## (a) (b) <-classified as
## ---- ----
## 1414 1328 (a): class 0
## 956 1786 (b): class 1
##
##
## Attribute usage:
##
## 100.00% allorder_amount
## 36.36% client_version
## 14.19% shipping_state
##
##
## Time: 0.0 secs