SLIDE 1
1
2
Survival Analysis Node
2
Manchester UK Professor Jorge Ribeiro Patrick Ribeiro 1 SAS/ETS - - PowerPoint PPT Presentation
Manchester UK Professor Jorge Ribeiro Patrick Ribeiro 1 SAS/ETS - - PowerPoint PPT Presentation
Survival Data Mining using Enterprise Miner and Proportional Hazard Cox Model 25 th June 2015 Manchester UK Professor Jorge Ribeiro Patrick Ribeiro 1 SAS/ETS Econometrics Time Series Enterprise Miner 13.2 PROC ARIMA Survival Analysis
SLIDE 2
SLIDE 3
3
Model 1 - Time to Next Purchase Survival Discrete Model
SLIDE 4
4
SLIDE 5
5
Survival Analysis Node
5
Enterprise Miner 13.2
SLIDE 6
6
1.1 - Model 1 - Time to Next Purchase Survival Discrete Model
SLIDE 7
7
1.2 - Model 1 - Time to Next Purchase
“People are much more likely to get on a bus if they know where it is going”.
Steps Plan
SLIDE 8
8
1.2 - Model 1 - Time to Next Purchase
SLIDE 9
9
1.2 - Model 1 - Time to Next Purchase
SLIDE 10
10 10
1.2 - Model 1 - Time to Next Purchase
SLIDE 11
11 11
Final Model - Hazard Function
SLIDE 12
12 12
Final Model - Benefit graph
SLIDE 13
13 13
Final Model
SLIDE 14
14 14
SLIDE 15
15 15
1.2 - Model 1 - Time to Next Purchase
SLIDE 16
16
PROC ARIMA / PROC AUTOREG
16
SAS/ETS – Econometrics Time Series
SLIDE 17
17 17
SAS/ETS – Econometrics Time Series
The Cross-Correlation Function
t t
H L
4
Jan
t t
H L
1 t t
H L
2
Oct Nov Dec Jan Feb Lag Apr Dec Jan Feb Mar Dec
SLIDE 18
18
PROC ARIMA / PROC AUTOREG
18
SAS/ETS – Econometrics Time Series
Primary Event Variables Royal Wedding Bank Holiday Price Marketing Campaign Point/Pulse Step Ramp tevent
SLIDE 19
19
PROC ARIMA / PROC AUTOREG
19
SAS/ETS – Econometrics Time Series
SLIDE 20
20
Simulation Studio 13.2
20
SAS/OR – Operational Research
SLIDE 21
21
Simulation Studio 13.2
21
SLIDE 22
22 22
2 - Model 2 - Call Centre Demand Call Centre Demand Model
SLIDE 23
23 23
2.1 - Model 2 – Call Centre
Wait Time Goal = 30 Wait Time Max = 90
SLIDE 24
24 24
2.2 - Model 2 – Call Centre
Wait Time Goal = 30 Wait Time Max = 90
SLIDE 25
25 25
2.3 - Model 2 – Call Centre
Wait Time Goal = 30 Wait Time Max = 90
SLIDE 26
26 26
2.4 - Model 2 – Call Centre
Wait Time Goal = 30 Wait Time Max = 90
SLIDE 27
27 27
2.5 - Model 2 – Call Centre
SLIDE 28
28 28
3.1 - Model 3 – Stress Test and Scenario Analysis
SLIDE 29
29 29
SLIDE 30
30 30
62 days for data preparation 6 days for modelling
SLIDE 31
31
SLIDE 32
32
Step 1 – Economic variables
Economic Variables Unemployment GDP Inflation Cash rate Credit availability House prices Commercial property prices Commodity prices Swap rates Equity prices
SLIDE 33
33
Cox Proportional Hazards Model
1 1
{ ... }
( ) ( )
i k ik
X X i
h t e h t
Baseline Hazard function - involves time but not predictor variables Linear function of a set of predictor variables - does not involve time
...
SLIDE 34
34
Step 3 – Model
PROC PHREG DATA = MODEL COVSANDWICH(AGGREGATE); CLASS Risk ; MODEL (START,END)*DEFAULT(0) = Risk P1GDP UNEMPLOYMENT; ID CUSTOMER_ID; HAZARDRATIO Risk / DIFF=REF; HAZARDRATIO P1GDP / UNITS = 1 2 3 5; HAZARDRATIO UNEMPLOYMENT / UNITS = 1 2 3 5; RUN;
PD_Band Risk 1 to 5 1 6 to 11 5 12 to 16 09 17 to 18 12 19 to 20 15
SLIDE 35
35
SAS Results
For each 1 unit increase in the GDP, the Hazard of Default goes down by an estimated 16.7 %.
0.18257
e 0.833
100*(0.833 1) 16.7%
SLIDE 36
36
SAS Results
For each 1 unit increase in the Unemployment, the Hazard of Default increases by an estimated 25.5 %.
0.22684
e 1.255 100*(1.255 1) 25.5%
Risk
SLIDE 37
37
SAS Results A customer in the Band 01 has a ONLY 8.7%
the risk of Default (or - 91.3%) compared to a customer in the Band 15 (the reference Band).
2.44279
e 0.087
100*(0.087 1) 91.3%
HAZARD RATIO (BAND 01) 0.087 HAZARD RATIO (BAND 15)
SLIDE 38
38
100*(0.087 1) 91.3%
HAZARD RATIO (BAND 01) 0.087 HAZARD RATIO (BAND 15)
HAZARDRATIO Risk / DIFF=REF;
SAS Results A customer in the Band 01 has a ONLY 8.7%
the risk of Default (or - 91.3%) compared to a customer in the Band 15 (the reference Band).
Risk
SLIDE 39
39
SAS Results
Output 7
PROC PHREG DATA = MODEL COVSANDWICH(AGGREGATE); CLASS Risk (PARAM=REF REF='15') ; MODEL (START,END)*DEFAULT(0) = Risk P1GDP UNEMPLOYMENT; ID CUSTOMER_ID; HAZARDRATIO P1GDP / UNITS = 1 2 3 5; HAZARDRATIO UNEMPLOYMENT / UNITS = 1 2 3 5; RUN;
100*(0.694 1) 30.6% 100*(0.578 1) 42.2% 100*(0.401 1) 59.9%
SLIDE 40
40
SAS Results
Output 8
PROC PHREG DATA = MODEL COVSANDWICH(AGGREGATE); CLASS Risk (PARAM=REF REF='15'); MODEL (START,END)* DEFAULT(0) = Risk P1GDP UNEMPLOYMENT; ID CUSTOMER_ID; HAZARDRATIO P1GDP / UNITS = 1 2 3 5; HAZARDRATIO UNEMPLOYMENT / UNITS = 1 2 3 5; RUN;
100*(1.574 1) 57.4% 100*(1.975 1) 97.5% 100*(3.109 1) 210.9%
SLIDE 41
41
Survival Function
Scenario Analysis 1 P1GDP=1.1; Unemployment=6; Scenario Analysis 2 P1GDP=0.8; Unemployment=10;
SLIDE 42
42
Forecast under Scenario
SLIDE 43
43 43
Go Further
Introduction to Survival Analysis using PH Cox Models Applying Survival Analysis for Business
SLIDE 44
44 44
Go Further
Survival Data Mining Programming Approach Survival Data Mining Using Enterprise Miner
SLIDE 45
45 45
Go Further – Books
SLIDE 46
46 46
Go Further – Books
SLIDE 47
47
Questions
47
www.modellingtraining.com info@modellingtraining.com
- SAS code
- Results