Downloaded 636 times
![Information value
%Good %Bad
[x] [Y]
< 5 1850 150 29% 5% 6.31 0.25 0.80 0.20
5-30 1600 400 25% 12% 2.05 0.13 0.31 0.04
31 - 60 1200 600 19% 18% 1.02 0.00 0.01 0.00
60 - 90 900 900 14% 28% 0.51 (0.14) -0.29 0.04
>= 91 800 1200 13% 37% 0.34 (0.24) -0.47 0.11
Total 6350 3250 0.39
IV
Utilizatio
n %
# of
Good
# of Bad
%Good /
%Bad
X -Y
WOE =
Log (X/Y)](https://image.slidesharecdn.com/introductiontopredictivemodelingv1-130520024629-phpapp02/85/Introduction-to-predictive-modeling-v1-28-320.jpg)
![Information value
%Good %Bad
[x] [Y]
< 5 1850 150 29% 5% 6.31 0.25 0.80 0.20
5-30 1600 400 25% 12% 2.05 0.13 0.31 0.04
31 - 60 1200 600 19% 18% 1.02 0.00 0.01 0.00
60 - 90 900 900 14% 28% 0.51 (0.14) -0.29 0.04
>= 91 800 1200 13% 37% 0.34 (0.24) -0.47 0.11
Total 6350 3250 0.39
IV
Utilizatio
n %
# of
Good
# of Bad
%Good /
%Bad
X -Y
WOE =
Log (X/Y)](https://image.slidesharecdn.com/introductiontopredictivemodelingv1-130520024629-phpapp02/75/Introduction-to-predictive-modeling-v1-28-2048.jpg)
Uploaded byVenkata Reddy Konasani
Introduction to predictive modeling v1
The document introduces predictive modeling, explaining its application in various business sectors, especially in credit risk assessment. It describes how banks use historical data to predict loan application outcomes, employing a statistical model to forecast customer behavior and potential defaults. The document also outlines the processes involved in model building, including data analysis and variable selection, with real-world examples illustrating the importance of predictive analytics.
More Related Content
Similar to Introduction to predictive modeling v1
Introduction to predictive modeling v1
- 1.
- 2. Contents • Fancy visualizationsto Predictive Modeling • The Business Problem • What is Predictive Modeling • The Horse Race Analogy • Credit Risk Model Building • Other Applications of Predictive Modeling
- 3.
- 4. DataVisualizations Data visualizations canhelp you make the right decisions quickly… that will ultimately lead to success.
- 5. Data Analysis &Predictive Modeling Increasingly, business rely on intelligent tools and techniques to analyze data systematically to improve decision-making. Retail sales analytics Financial services analytics Telecommunications SupplyChain analytics Transportation analytics Risk & Credit analytics Talent analytics Marketing analytics Behavioral analytics Collections analytics Fraud analytics Pricing analytics
- 6. Dude we allknow that Data analytics rocks & Predictive modeling is going to rule!!!
- 7. • Credit Cards& Loans- How does bank approve or decline CC and Loan applications • Citi Bank : Present in more than 90 countries. • More than 100,000 customers apply for credit cards/loans every month • How do they approve or reject credit card applications? Business Problem
- 8. How does bankapprove or decline CC and Loan applications? • Approve if the applicant is an auto driver in Bangalore • Approve if the applicant’s boy friend / girl friend is rich • Approve if the applicant is a u-19 cricketer • Decline if the applicant is a software professional from Bangalore Any suggestions… Chill maddi
- 9. The Problem Who willrun away with my money? • Citi Bank : Present in more than 90 countries. • More than 100,000 customers apply for credit cards/loans every month • All of them have different characteristics • Out of 10o,000 customers, who all have the higher probability of default/ Charge off? • Basically, who will run away with my money? • We need to predict the probability of “Running away”
- 10. How to predict? Wehave historical data and we have statistics
- 11. Bank builds amodel that gives a score to each customer “Developing set of equations or mathematical formulation to forecast future behaviors based on current or historical data.” Applicants
- 12. Predictive Modeling Lets tryto understand predictive modeling Predictive Modeling – Fitting a model to the data to predict the future. • Predicting the future –and it is so easy some times • Who is going to score more runs in IPL-2014? • That’s it you predicted the future.. • BTW how did you predict? • Predicting the future based on historical data is nothing but Predictive modeling
- 13. Predictive Modeling Lets tryto understand predictive modeling Predictive Modeling – Fitting a model to the data to predict the future. • Who is going to score more runs in IPL 2014? • Predicting the future …well it is not that easy …
- 14. Predictive Modeling Horse RaceAnalogy Howto bet on best horse in a horse race
- 15. The Historical Data Winvs. Loss record in past 2 years • Long legs: 75% (Horses with long legs won 75% of the times) • BreedA: 55%, Breed B: 15 % Others : 30% • T/L (Tummy to length) ratio <1/2 :75 % • Gender: Male -68% • Head size: Small 10%, Medium 15% Large 75% • Country: Africa -65%
- 16. Given the historicaldata Which one of these two horses would you bet on? Kalyan Chethak Length of legs 150 cm 110 cm Breed A F T/L ratio 0.3 0.6 Gender Male Female Head size Large Small Country India India
- 17. Given the historicaldata Which one would you bet on….now? Kalyan Chethak Length of legs 110 cm 150 cm Breed C A T/L ratio 0.45 0.60 Gender Male Female Head size Small Large Country Africa India
- 18. Given the historicaldata What about best one in this lot? Horse-1 Horse-2 Horse-3 Horse-4 Horse-5 Horse-6 Horse-7 Horse-8 Horse-9 Horse-10 Length of legs 109 114 134 130 149 120 104 117 115 135 Breed C A B A F K L B C A T/L ratio 0.1 0.8 0.5 1.0 0.3 0.3 0.3 0.6 0.7 0.9 Gender Male Female Male Female Male Female Male Female Male Female Head size L S M M L L S M L M Country Africa India Aus NZ Africa Africa India India Aus Africa
- 19. Citi has asimilar problem? Who is going to run away with my money? • Given Historical of the customers we want to predict the probability of bad • We have the data of each customer on • Customer previous loans, customer previous payments, length of account credit history, other credit cards and loans, job type, income band etc., • We want to predict the probability of default
- 20. Credit Risk ModelBuilding-Four main steps 1. Study historical data • What are the causes(Customer Characteristics) • What are the effects(Charge off) 2. Identify the most impacting factors 3. Find the exact impact of each factor(Quantify it) 4. Use these coefficients for future
- 21. The Historical Dataof Customers • Contains all the information about customers • Contains information across more than 500 variables • Portion of data is present in the application form • Portion of the data is available with bank • Lot of data is maintained in bureau • Social Security number –in US • PAN Number in India Attribute Value SSN 111259005 Age 27 Number of dependents 2 Number of current loans 1 Number of credit cards 1 Number Installments 30days late in last 2 years 4 Average utilization % in last 2 years 30% Time since accounts opened 60 months Number of previous applications for credit card 2 Bankrupt NO
- 22. The Historical Dataof Customers:Types of variables Payment History • Account payment information on specific types of accounts (credit cards, retail accounts, installment loans, finance company accounts, mortgage, etc.) • Presence of adverse public records (bankruptcy, judgements, suits, liens, wage attachments, etc.), collection items, and/or delinquency (past due items) • Severity of delinquency (how long past due) • Amount past due on delinquent accounts or collection items • Time since (recency of) past due items (delinquency), adverse public records (if any), or collection items (if any) • Number of past due items on file • Number of accounts paid as agreed
- 23. More variables AmountsOwed • Amountowing on accounts • Amount owing on specific types of accounts • Lack of a specific type of balance, in some cases • Number of accounts with balances • Proportion of credit lines used (proportion of balances to total credit limits on certain types of revolving accounts) • Proportion of installment loan amounts still owing (proportion of balance to original loan amount on certain types of installment loans) Length of Credit History • Time since accounts opened • Time since accounts opened, by specific type of account • Time since account activity New Credit • Number of recently opened accounts, and proportion of accounts that are recently opened, by type of account • Number of recent credit inquiries • Time since recent account opening(s), by type of account • Time since credit inquiry(s) • Re-establishment of positive credit history following past payment problems
- 24.
- 25. Identifying most impactingfactors Out of 500 what are those 20 important attributes
- 26. Selecting the variables-How do you do that? Num_Enq Good Bad 0 6400 5 1 5800 69 2 5445 134 3 4500 250 4 4070 375 5 3726 470 6 2879 650 7 1893 876 8 1236 987 9 354 1298 36303 5114 Num_companies Good Bad 0 4858 594 1 4291 602 2 4410 577 3 4738 529 4 4506 707 5 4976 796 6 4723 770 7 4407 583 8 4157 825 9 4970 964 46036 6947
- 27. Bad Rate Num_Enq GoodBad 0 6400 5 1 5800 69 2 5445 134 3 4500 250 4 4070 375 5 3726 470 6 2879 650 7 1893 876 8 1236 987 9 354 1298 36303 5114 Bad Rate 0% 1% 3% 5% 7% 9% 13% 17% 19% 25% Num_companies Good Bad 0 4858 594 1 4291 602 2 4410 577 3 4738 529 4 4506 707 5 4976 796 6 4723 770 7 4407 583 8 4157 825 9 4970 964 46036 6947 Bad Rate 9% 9% 8% 8% 10% 11% 11% 8% 12% 14%
- 28. Information value %Good %Bad [x][Y] < 5 1850 150 29% 5% 6.31 0.25 0.80 0.20 5-30 1600 400 25% 12% 2.05 0.13 0.31 0.04 31 - 60 1200 600 19% 18% 1.02 0.00 0.01 0.00 60 - 90 900 900 14% 28% 0.51 (0.14) -0.29 0.04 >= 91 800 1200 13% 37% 0.34 (0.24) -0.47 0.11 Total 6350 3250 0.39 IV Utilizatio n % # of Good # of Bad %Good / %Bad X -Y WOE = Log (X/Y)
- 30. Model Building Historical Data Themodel equation
- 31. Model Building logistic regressionmodel to predict the probability of default • Probability of bad = w1(Var1)+w2(Var2) +w3(Var1)……+w20(var20) • Logistic regression gives us those weights • Predicting the probability • Odds of bad=0.13(number of cards)+ 0.21 (utilization)+…….+0.06(number of loan applications) • That’s it ….we are done
- 32. Credit Risk ModelBuilding-Example Attributes used on the model 1. Age Application form 2. MonthlyIncome1Application form 3. Dependents Application form 4. Number of times 30DPD in last two years Bureau 5. Debt Ratio Bureau /Bank 6. Utilization Bureau /Bank 7. Number of loans Bureau /Bank
- 33. Model building andImplementation Data: http://www.kaggle.com/c/GiveMeSomeCredit Model building script: The final Model
- 34. Actual model BuildingSteps Objective & Portfolio identification Customer score vs. account score Exclusions Observation point and/or window Performance window Bad definition Segmentation Validation plan and samples Variable Selection Model Building Scoring and Validation Score Implementation
- 35. Marketing Example Predicting theresponse probability to a marketing Campaign • Selling Mobile phones – Marketing campaign • Who should we target? • Consider historical data of mobile phone buyers • See their characteristics • Find top impacting characteristics • Find weight of each characteristic • Score new population • Decide on the cut off • Try to sell people who score more than cut off
- 36. Other Applications ofModel Building • Fraud transactions scorecard – Fraud identification based on attributes like transaction amount, place, time, frequency of transactions etc., • Attrition modeling – Predicting employee attrition based on their characteristics
- 37.