金融风控相关业务介绍
学习目标
-
知道常见信贷风险
-
知道机器学习风控模型的优势
-
知道信贷领域常用术语含义
1 信贷&风控介绍
2 常见信贷产品及常见风险介绍
-
信贷产品介绍
个人信贷产品 | | |
---|
大额借贷 | 房贷车贷 | |
| 小微企业贷 | |
小额借贷 | 消费贷 | 蚂蚁花呗,京东白条 |
| 现金贷 | 蚂蚁借呗,京东金条,微粒贷,各类网贷 |
数据服务 | 信用分服务 | 芝麻信用分,京东小白分 |
| 信用数据服务 | 同盾数据,百融,集奥,大峰... |
额度 | 500~3000 |
---|
利率 | 24%~36% |
期限 | 714,30天 |
放款形式 | 借给现金,不限场景 |
可选功能 | 订单展期 |
现金贷产品 | 年化利率 | 现金贷产品 | 年化利率 |
---|
苏宁金融 | 24% | 国美易卡 | 34% |
蚂蚁借呗 | 24% | 马上消费 | 35% |
微粒贷 | 24% | 招联金融 | 36% |
有钱花 | 24% | 桔子分期 | 36% |
京东金条 | 24% | 拍拍贷 | 36% |
360借条 | 24% | 趣店 | 36% |
小米金融 | 24% | 捷信 | 36% |
美团生活费 | 24% | 宜人贷 | 44% |
分期乐 | 24% | 玖富 | 50% |
-
消费贷
信用卡,花呗,白条等产品,有账单日,还款日
申请消费贷 -> 额度授信->客户使用消费贷消费
额度 | 1000~10000 |
---|
利率 | 24% |
账期 | 30天 |
放款形式 | 指定消费场景 |
可选功能 | 最低还款,账单展期,账单分期,停息挂账,临时额度,备用金 |
-
产品类型:
-
还款方式:
-
砍头息:短期产品, (服务费)
-
等额本金
-
等额本息
-
常见风险
-
冒名顶替,黑产骗贷
-
多头借贷,借新还旧
-
POS机套现,以少换多
-
针对风控模型,制作数据
3 风控相关术语介绍
| 解释 |
---|
DPD | Day past due 逾期天数 DPD0为到期当日,DPD1为逾期一日,DPD7为逾期一周 |
FPD | First time past due 首次逾期天数 |
F/S/T/QPD | 首次 二次 三次 四次 逾期天数 |
M1 | 逾期 [1, 30)天 M 是英文“Months”的首写字母 |
M1+ | 逾期[30, inf]天 |
default | 坏账 |
delinquency | 拖欠 |
flow rate | 流动率 一般指M1向M2,M2向M3转移的比例 |
bad rate | 坏账率 当月不良资产数/总资产数 |
vintage | 账龄分析 |
4 风控业务案例
4.1 案例背景介绍
-
通过对业务数据分析了解信贷业务状况
-
数据集说明
-
从开源数据改造而来,基本反映真实业务数据
-
销售,客服可以忽略
-
账单周期,放款日期
-
账单金额-实收金额 = 未收金额
-
应付日期为还款时间
-
账期分成两种 60天和90天
-
实际到账日为空白 说明没还钱
-
-
通过对贷后业务数据的分析要分析出如下内容
-
每个季度账单金额和坏账率(逾期90天以上)
-
所有未收金额/所有账单金额
-
未收金额 = 账单金额-实收金额
-
每个季度60天账期 入催率,90天账单 入催率
-
不同逾期天数的回款情况
-
历史逾期天数 有逾期 已经还完了
-
当前逾期天数 现在还欠着钱,没还完
4.2 代码实现
<span style="background-color:#f8f8f8"><span style="color:#333333"><span style="color:#770088">import</span> <span style="color:#000000">pandas</span> <span style="color:#770088">as</span> <span style="color:#000000">pd</span>
<span style="color:#770088">import</span> <span style="color:#000000">datetime</span>
<span style="color:#770088">from</span> <span style="color:#000000">pyecharts</span>.<span style="color:#000000">charts</span> <span style="color:#770088">import</span> <span style="color:#981a1a">*</span>
<span style="color:#770088">from</span> <span style="color:#000000">pyecharts</span> <span style="color:#770088">import</span> <span style="color:#000000">options</span> <span style="color:#770088">as</span> <span style="color:#000000">opts</span>
<span style="color:#000000">df1</span> = <span style="color:#000000">pd</span>.<span style="color:#000000">read_excel</span>(<span style="color:#aa1111">'data/业务数据.xls'</span>)
<span style="color:#aa5500">#要使用原始数据构建新指标,所以保留原始数据,copy新的数据,在新的数据中创建新指标</span>
<span style="color:#000000">df2</span> = <span style="color:#000000">df1</span>.<span style="color:#000000">copy</span>()
<span style="color:#000000">df2</span>.<span style="color:#000000">head</span>()</span></span>
显示结果
| 销售 | 账单状态 | 账单周期 | 账单金额 | 开票金额 | 实收金额 | 未收金额 | 预计付款日 | 应付日期 | 商务催收日期 | 账期 | 实际到账日 | 开票日期 | 客服 |
---|
0 | s101 | 未确认 | 2019-05 | 29805.0 | NaN | NaN | NaN | 2019-07-31 | 2019-07-31 | 2019-08-15 | 60 | NaN | NaN | a201 |
1 | s102 | 未确认 | 2019-05 | 1572.6 | NaN | NaN | NaN | 2019-07-31 | 2019-07-31 | 2019-08-15 | 60 | NaN | NaN | a202 |
2 | s103 | 已确认 | 2019-04 | 487551.2 | 487551.2 | NaN | 487551.2 | 2019-06-30 | 2019-06-30 | 2019-07-15 | 60 | NaN | 05-16 | a203 |
3 | s104 | 已确认 | 2019-04 | 378835.0 | 378835.0 | NaN | 378835.0 | 2019-07-31 | 2019-07-31 | 2019-08-15 | 90 | NaN | 05-08 | a204 |
4 | s105 | 已确认 | 2019-04 | 326866.0 | 326866.0 | NaN | 326866.0 | 2019-07-31 | 2019-07-31 | 2019-08-15 | 90 | NaN | 05-10 | a205 |
<span style="background-color:#f8f8f8"><span style="color:#333333"><span style="color:#000000">df2</span>.<span style="color:#000000">info</span>()</span></span>
显示结果
<span style="background-color:#f8f8f8"><class <span style="color:#aa1111">'pandas.core.frame.DataFrame'</span>>
RangeIndex: <span style="color:#116644">5257</span> entries, <span style="color:#116644">0</span> to <span style="color:#116644">5256</span>
Data columns (total <span style="color:#116644">14</span> columns):
<span style="color:#aa5500"># Column Non-Null Count Dtype </span>
<span style="color:#0000cc">---</span> <span style="color:#0000cc">------</span> <span style="color:#0000cc">--------------</span> <span style="color:#0000cc">-----</span>
<span style="color:#116644">0</span> 销售 <span style="color:#116644">5257</span> non-null object
<span style="color:#116644">1</span> 账单状态 <span style="color:#116644">5257</span> non-null object
<span style="color:#116644">2</span> 账单周期 <span style="color:#116644">5257</span> non-null object
<span style="color:#116644">3</span> 账单金额 <span style="color:#116644">5257</span> non-null float64
<span style="color:#116644">4</span> 开票金额 <span style="color:#116644">5010</span> non-null float64
<span style="color:#116644">5</span> 实收金额 <span style="color:#116644">4470</span> non-null float64
<span style="color:#116644">6</span> 未收金额 <span style="color:#116644">5010</span> non-null float64
<span style="color:#116644">7</span> 预计付款日 <span style="color:#116644">5256</span> non-null object
<span style="color:#116644">8</span> 应付日期 <span style="color:#116644">5257</span> non-null object
<span style="color:#116644">9</span> 商务催收日期 <span style="color:#116644">5257</span> non-null object
<span style="color:#116644">10</span> 账期 <span style="color:#116644">5257</span> non-null int64
<span style="color:#116644">11</span> 实际到账日 <span style="color:#116644">4387</span> non-null object
<span style="color:#116644">12</span> 开票日期 <span style="color:#116644">4996</span> non-null object
<span style="color:#116644">13</span> 客服 <span style="color:#116644">5257</span> non-null object
dtypes: float64(4), int64(1), object(9)
memory usage: <span style="color:#116644">575</span>.1<span style="color:#981a1a">+</span> KB</span>
<span style="background-color:#f8f8f8"><span style="color:#333333"><span style="color:#000000">df2</span>.<span style="color:#000000">describe</span>()</span></span>
显示结果
| 账单金额 | 开票金额 | 实收金额 | 未收金额 | 账期 |
---|
count | 5.257000e+03 | 5.010000e+03 | 4.470000e+03 | 5.010000e+03 | 5257.000000 |
mean | 4.073241e+04 | 4.096896e+04 | 4.082419e+04 | 4.684636e+03 | 64.539661 |
std | 8.176172e+04 | 8.007245e+04 | 7.970628e+04 | 2.888464e+04 | 15.622765 |
min | 0.000000e+00 | 2.500000e+01 | 0.000000e+00 | 0.000000e+00 | 0.000000 |
25% | 5.103000e+03 | 5.300000e+03 | 5.112250e+03 | 0.000000e+00 | 60.000000 |
50% | 1.436500e+04 | 1.486560e+04 | 1.434000e+04 | 0.000000e+00 | 60.000000 |
75% | 4.178000e+04 | 4.220250e+04 | 4.170750e+04 | 0.000000e+00 | 75.000000 |
max | 1.508796e+06 | 1.356215e+06 | 1.301665e+06 | 1.277098e+06 | 90.000000 |
<span style="background-color:#f8f8f8"><span style="color:#333333"><span style="color:#aa5500"># 获取最大的日期,作为当前时间</span>
<span style="color:#000000">today_time</span> = <span style="color:#000000">pd</span>.<span style="color:#000000">to_datetime</span>(<span style="color:#000000">df2</span>.<span style="color:#000000">实际到账日</span>.<span style="color:#000000">fillna</span>(<span style="color:#aa1111">'0'</span>).<span style="color:#000000">max</span>())
<span style="color:#000000">df2</span>[<span style="color:#aa1111">'实收金额'</span>] = <span style="color:#000000">df2</span>.<span style="color:#000000">实收金额</span>.<span style="color:#000000">fillna</span>(<span style="color:#116644">0</span>)
<span style="color:#000000">df2</span>[<span style="color:#aa1111">'开票金额'</span>] = <span style="color:#000000">df2</span>.<span style="color:#000000">开票金额</span>.<span style="color:#000000">fillna</span>(<span style="color:#116644">0</span>)
<span style="color:#000000">df2</span>[<span style="color:#aa1111">'未收金额'</span>] = <span style="color:#000000">df2</span>.<span style="color:#000000">未收金额</span>.<span style="color:#000000">fillna</span>(<span style="color:#116644">0</span>)
<span style="color:#000000">df2</span>[<span style="color:#aa1111">'账单周期'</span>] = <span style="color:#000000">pd</span>.<span style="color:#000000">to_datetime</span>(<span style="color:#000000">df2</span>.<span style="color:#000000">账单周期</span>)
<span style="color:#000000">df2</span>[<span style="color:#aa1111">'应付日期'</span>] = <span style="color:#000000">pd</span>.<span style="color:#000000">to_datetime</span>(<span style="color:#000000">df2</span>.<span style="color:#000000">应付日期</span>)
<span style="color:#000000">df2</span>[<span style="color:#aa1111">'实际到账日'</span>] = <span style="color:#000000">pd</span>.<span style="color:#000000">to_datetime</span>(<span style="color:#000000">df2</span>.<span style="color:#000000">实际到账日</span>).<span style="color:#000000">fillna</span>(<span style="color:#000000">today_time</span>)</span></span>
<span style="background-color:#f8f8f8"><span style="color:#333333"><span style="color:#000000">df2</span>[<span style="color:#aa1111">'是否到期'</span>] = <span style="color:#000000">df2</span>.<span style="color:#000000">apply</span>(<span style="color:#770088">lambda</span> <span style="color:#000000">x</span> : <span style="color:#116644">0</span> <span style="color:#770088">if</span> <span style="color:#000000">x</span>.<span style="color:#000000">应付日期</span> <span style="color:#981a1a">></span> <span style="color:#000000">today_time</span> <span style="color:#770088">else</span> <span style="color:#116644">1</span>,<span style="color:#000000">axis</span>=<span style="color:#116644">1</span>)
<span style="color:#000000">df2</span>[<span style="color:#aa1111">'是否到期90天'</span>] = ( <span style="color:#000000">today_time</span> <span style="color:#981a1a">-</span> <span style="color:#000000">df2</span>.<span style="color:#000000">应付日期</span> ).<span style="color:#000000">map</span>(<span style="color:#770088">lambda</span> <span style="color:#000000">x</span> : <span style="color:#116644">1</span> <span style="color:#770088">if</span> <span style="color:#000000">x</span>.<span style="color:#000000">days</span> <span style="color:#981a1a">></span>= <span style="color:#116644">90</span> <span style="color:#770088">else</span> <span style="color:#116644">0</span>)
<span style="color:#000000">df2</span>[<span style="color:#aa1111">'未收金额2'</span>] = (<span style="color:#000000">df2</span>.<span style="color:#000000">账单金额</span> <span style="color:#981a1a">-</span> <span style="color:#000000">df2</span>.<span style="color:#000000">实收金额</span>)
<span style="color:#000000">df2</span>[<span style="color:#aa1111">'历史逾期天数'</span>] = <span style="color:#000000">df2</span>.<span style="color:#000000">apply</span>(<span style="color:#770088">lambda</span> <span style="color:#000000">x</span> : (<span style="color:#000000">x</span>.<span style="color:#000000">实际到账日</span> <span style="color:#981a1a">-</span> <span style="color:#000000">x</span>.<span style="color:#000000">应付日期</span>).<span style="color:#000000">days</span> <span style="color:#770088">if</span> <span style="color:#000000">x</span>.<span style="color:#000000">未收金额2</span> == <span style="color:#116644">0</span> <span style="color:#770088">else</span> (<span style="color:#000000">today_time</span> <span style="color:#981a1a">-</span> <span style="color:#000000">x</span>.<span style="color:#000000">应付日期</span>).<span style="color:#000000">days</span>,<span style="color:#000000">axis</span>=<span style="color:#116644">1</span>)
<span style="color:#000000">df2</span>[<span style="color:#aa1111">'当前逾期天数'</span>] = <span style="color:#000000">df2</span>.<span style="color:#000000">apply</span>(<span style="color:#770088">lambda</span> <span style="color:#000000">x</span> : (<span style="color:#000000">x</span>.<span style="color:#000000">历史逾期天数</span>) <span style="color:#770088">if</span> <span style="color:#000000">x</span>.<span style="color:#000000">未收金额2</span> <span style="color:#981a1a">></span> <span style="color:#116644">0</span> <span style="color:#770088">else</span> <span style="color:#116644">0</span> ,<span style="color:#000000">axis</span> = <span style="color:#116644">1</span>) </span></span>
<span style="background-color:#f8f8f8"><span style="color:#333333"><span style="color:#000000">df3</span> =<span style="color:#000000">df2</span>.<span style="color:#000000">copy</span>()
<span style="color:#aa5500">#创建’账单季度‘字段,将日期转换成季度</span>
<span style="color:#000000">df3</span>[<span style="color:#aa1111">'账单季度'</span>] = <span style="color:#000000">df3</span>[<span style="color:#aa1111">'账单周期'</span>].<span style="color:#000000">map</span>(<span style="color:#770088">lambda</span> <span style="color:#000000">x</span> : <span style="color:#000000">x</span>.<span style="color:#000000">to_period</span>(<span style="color:#aa1111">'Q'</span>))
<span style="color:#aa5500">#提取2017年3季度到2018年4季度数据</span>
<span style="color:#000000">df3</span> = <span style="color:#000000">df3</span>[(<span style="color:#000000">df3</span>[<span style="color:#aa1111">'账单季度'</span>]<span style="color:#981a1a"><</span>=<span style="color:#aa1111">'2018Q4'</span>) <span style="color:#981a1a">&</span> (<span style="color:#000000">df3</span>[<span style="color:#aa1111">'账单季度'</span>]<span style="color:#981a1a">></span>=<span style="color:#aa1111">'2017Q3'</span>)]
<span style="color:#000000">df3</span>.<span style="color:#000000">shape</span></span></span>
显示结果
<span style="background-color:#f8f8f8">(3856, 21)</span>
<span style="background-color:#f8f8f8"><span style="color:#333333">#账单金额
fn1 = df3.groupby('账单季度')[['账单金额']].sum()
fn1.columns = ['账单金额']
fn1
</span></span>
显示结果
| 账单金额 |
---|
账单季度 | |
2017Q3 | 8247952.62 |
2017Q4 | 11643604.99 |
2018Q1 | 17149674.79 |
2018Q2 | 31097661.29 |
2018Q3 | 38292071.12 |
2018Q4 | 51963089.64 |
<span style="background-color:#f8f8f8"><span style="color:#333333">#90天到期金额
df4 = df3[(df3.是否到期90天 == 1)]
fn2 = df4.groupby('账单季度')[['账单金额']].sum()
fn2.columns = ['到期金额']
fn2
</span></span>
显示结果
| 到期金额 |
---|
账单季度 | |
2017Q3 | 8247952.62 |
2017Q4 | 11643604.99 |
2018Q1 | 17149674.79 |
2018Q2 | 31097661.29 |
2018Q3 | 38292071.12 |
2018Q4 | 28265677.59 |
<span style="background-color:#f8f8f8"><span style="color:#333333">df4 = df3[(df3.是否到期90天 == 1)]
fn3 = df4.groupby('账单季度')[['未收金额2']].sum()
fn3.columns = ['当前逾期90+金额']
fn3
</span></span>
显示结果
| 当前逾期90+金额 |
---|
账单季度 | |
2017Q3 | 63883.0 |
2017Q4 | 57380.0 |
2018Q1 | 64283.0 |
2018Q2 | 106930.0 |
2018Q3 | 412920.1 |
2018Q4 | 304183.0 |
<span style="background-color:#f8f8f8"><span style="color:#333333">dfs = [fn1,fn2,fn3]
final1 = pd.concat(dfs,axis=1)
final1
</span></span>
显示结果
| 账单金额 | 到期金额 | 当前逾期90+金额 |
---|
账单季度 | | | |
2017Q3 | 8247952.62 | 8247952.62 | 63883.0 |
2017Q4 | 11643604.99 | 11643604.99 | 57380.0 |
2018Q1 | 17149674.79 | 17149674.79 | 64283.0 |
2018Q2 | 31097661.29 | 31097661.29 | 106930.0 |
2018Q3 | 38292071.12 | 38292071.12 | 412920.1 |
2018Q4 | 51963089.64 | 28265677.59 | 304183.0 |
<span style="background-color:#f8f8f8"><span style="color:#333333">final1['90+净坏账率'] = round(final1['当前逾期90+金额'] / final1.到期金额,3)
final1
</span></span>
显示结果
| 账单金额 | 到期金额 | 当前逾期90+金额 | 90+净坏账率 |
---|
账单季度 | | | | |
2017Q3 | 8247952.62 | 8247952.62 | 63883.0 | 0.008 |
2017Q4 | 11643604.99 | 11643604.99 | 57380.0 | 0.005 |
2018Q1 | 17149674.79 | 17149674.79 | 64283.0 | 0.004 |
2018Q2 | 31097661.29 | 31097661.29 | 106930.0 | 0.003 |
2018Q3 | 38292071.12 | 38292071.12 | 412920.1 | 0.011 |
2018Q4 | 51963089.64 | 28265677.59 | 304183.0 | 0.011 |
<span style="background-color:#f8f8f8"><span style="color:#333333">bar = (
Bar()
.add_xaxis(list(final1.index.values.astype(str)))
.add_yaxis(
"账单金额",
list(final1.账单金额),
yaxis_index=0,
color="#5793f3",
)
.set_global_opts(
title_opts=opts.TitleOpts(title="90+净坏账率"),
)
.extend_axis(
yaxis=opts.AxisOpts(
name="90+净坏账率",
type_="value",
min_=0,
max_=0.014,
position="right",
axisline_opts=opts.AxisLineOpts(
linestyle_opts=opts.LineStyleOpts(color="#d14a61")
),
axislabel_opts=opts.LabelOpts(formatter="{value}"),
)
)
)
line = (
Line()
.add_xaxis(list(final1.index.values.astype(str)))
.add_yaxis(
"90+净坏账率",
list(final1['90+净坏账率']),
yaxis_index=1,
color="#675bba",
label_opts=opts.LabelOpts(is_show=False),
)
)
bar.overlap(line).render_notebook()
</span></span>
显示结果
<span style="background-color:#f8f8f8"><span style="color:#333333">#60天账期的账单金额
df4 = df3[(df3.账期 == 60)&(df3.是否到期 == 1)]
fn1 = df4.groupby('账单季度')[['账单金额']].sum()
fn1.columns = ['60天账期的账单金额']
#60天账期的入催金额
df4 = df3[(df3.账期 == 60)&(df3.是否到期 == 1)&(df3.历史逾期天数>0)]
fn2 = df4.groupby('账单季度')[['未收金额2']].sum()
fn2.columns = ['60天账期的入催金额']
#90天账期的账单金额
df4 = df3[(df3.账期 == 90)&(df3.是否到期 == 1)]
fn3 = df4.groupby('账单季度')[['账单金额']].sum()
fn3.columns = ['90天账期的账单金额']
#90天账期的入催金额
df4 = df3[(df3.账期 == 90)&(df3.是否到期 == 1)&(df3.历史逾期天数>0)]
fn4 = df4.groupby('账单季度')[['未收金额2']].sum()
fn4.columns = ['90天账期的入催金额']
</span></span>
<span style="background-color:#f8f8f8"><span style="color:#333333">dfs = [fn1,fn2,fn3,fn4]
final2 = pd.concat(dfs,axis=1)
# final2 = fn1.merge(fn2,on='账单季度').merge(fn3,on='账单季度',how='left').merge(fn4,on='账单季度')
final2['60天账期入催率'] = round(final2['60天账期的入催金额'] / final2['60天账期的账单金额'],3)
final2['90天账期入催率'] = round(final2['90天账期的入催金额']/final2['90天账期的账单金额'],3)
final2
</span></span>
显示结果
| 60天账期的账单金额 | 60天账期的入催金额 | 90天账期的账单金额 | 90天账期的入催金额 | 60天账期入催率 | 90天账期入催率 |
---|
账单季度 | | | | | | |
2017Q3 | 4854770.94 | 36983.0 | 2769264.0 | 1900.0 | 0.008 | 0.001 |
2017Q4 | 6737327.99 | 52750.0 | 3921491.0 | 0.0 | 0.008 | 0.000 |
2018Q1 | 12106356.79 | 62460.0 | 4244304.0 | 800.0 | 0.005 | 0.000 |
2018Q2 | 19234086.87 | 13590.0 | 8427775.0 | 0.0 | 0.001 | 0.000 |
2018Q3 | 22830710.42 | 380265.1 | 9835629.0 | 8235.0 | 0.017 | 0.001 |
2018Q4 | 26337959.52 | 584789.5 | 17706430.0 | 325141.0 | 0.022 | 0.018 |
<span style="background-color:#f8f8f8"><span style="color:#333333">line = (
Line()
.add_xaxis(list(final1.index.values.astype(str)))
.add_yaxis(
"60天账期入催率",
list(final2['60天账期入催率']),
yaxis_index=0,
color="#675bba",
label_opts=opts.LabelOpts(is_show=False),
)
.set_global_opts(
title_opts=opts.TitleOpts(title="不同账期入催率"),
)
.add_xaxis(list(final1.index.values.astype(str)))
.add_yaxis(
"90天账期入催率",
list(final2['90天账期入催率']),
yaxis_index=0,
color="#d14a61",
label_opts=opts.LabelOpts(is_show=False),
)
)
line.render_notebook()
</span></span>
显示结果
<span style="background-color:#f8f8f8"><span style="color:#333333">df6 = df3[(df3.未收金额2 == 0)&(df3.是否到期 == 1)].copy()
#使用cut,讲数据按照逾期天数分箱,然后添加分箱之后结果标签
df6['历史逾期天数'] = pd.cut(df6['历史逾期天数'],bins=[-999,0,5,10,15,20,30,60,90,999],right=True,
labels=['0','1-5','6-10','11-15','16-20','21-30','31-60','61-90','91+'])
final3 = df6.groupby('历史逾期天数')[['账期']].count()
final3.columns = ['回收账单数']
final3
</span></span>
显示结果
| 回收账单数 |
---|
历史逾期天数 | |
0 | 2400 |
1-5 | 358 |
6-10 | 235 |
11-15 | 215 |
16-20 | 92 |
21-30 | 189 |
31-60 | 156 |
61-90 | 60 |
91+ | 88 |
<span style="background-color:#f8f8f8"><span style="color:#333333">ydata = final3['回收账单数'].values.tolist()
bar = (
Bar()
.add_xaxis(list(final3.index.values.tolist()))
.add_yaxis("收回账单数",ydata,yaxis_index=0,color="#675bba")
.set_global_opts(
title_opts=opts.TitleOpts(title="不同逾期天数的已收回账单数"),
)
)
bar.render_notebook()
</span></span>
显示结果
4.3 业务解读
-
从数据中看出,在2018年Q2季度之前,运营策略比较保守,坏账金额和入催率都比较低,
-
2018年Q2之后,有可能是由于运营策略调整,给更多的人放贷,但坏账率和入催率均在3%一下,在合理范围内
-
不同逾期天数收回账单的数据看,30天内能收回绝大部分账单
小结
-
知道常见信贷风险
-
知道机器学习风控模型的优势
-
批量,迅速,准确,同时处理大量贷款请求
-
在零售信贷业务场景下,与人工审核相比优势显著
-
知道信贷领域常用术语含义
-
DPD 逾期天数
-
FPD 首次逾期天数
-
bad rate 坏账率
-
M1 逾期一个月以内
清溪上: 您好,请问数据集可以在哪里获取呀
大大大大帅哥: 有具体的数据集吗
罗均 珞玘 AmosLuo: 请问这个课程叫什么名字,非常感兴趣,多谢博主啦
xiaobingbuhuitou: 大佬请问一下这边,在样本不均衡简介中提到的 “其中负样本所贡献的信息只有模型接收到的总信息的1/100”这个贡献度为样本数量倒数是如何计算的?
向量2019: 大神有code版本的吗?markdown看起来好麻烦