Sorry, you do not have access to this eBook
A subscription is required to access the full text content of this book.
Metrics
Views
28
In This Chapter
-
Credit Scoring Models
- The method of scoring
- The emergence of credit scoring
- Judgmental credit systems versus credit scoring systems
- Scoring models for corporate bankruptcy
- New developments
- Variable selection
- Score modeling, validation and follow-up
- Statistical methods
- Unsupervised learning
- Parametric supervised methods
- Non-parametric supervised methods
- k-nearest neighbor classifier
- Neural networks
- Neural networks as non-linear models
- Neural networks as a mathematical representation of biological systems
- Structure of the canonical neural network
- Loss function
- Learning rules
- Error backpropagation
- Examples
- Support vector machines
- Separating hyperplanes
- Hard margin classification
- Soft margin classification
- SVM regression
- Non-linear support vector machines
- Extension to the multi-class problem
- Model averaging
- Performance evaluation criteria and score consistency
- Exercises
- Elastic net regression
- Cross-validation of the ridge linear regression
- K-means and the Lloyd’s algorithm
- Derivation of the principal component analysis
- Two-class separation maximization
- Maximum likelihood estimation of the probit model
- Computation of feed-forward neural networks
- Primal and dual problems of support vector machines
- Derivation of the AdaBoost algorithm as the solution of the additive logit model
- Weighted estimation
- Conclusion
Abstract
Credit scoring refers to statistical models to measure the creditworthiness of a person or a company. They have progressively replaced judgemental systems and are now widely used by financial and banking institutions that check the credit rating and capacity of the borrower before to approve a loan. Therefore, credit scoring is at the heart of the decision-making system for granting credit. This is particularly true for consumer credit (mortgage, credit card, personal loan, etc.). Credit scoring models are also used for commercial firms, but their final outputs are generally not sufficient for making a decision. For instance, they can be completed with the knowledge of the relationship manager on the company.
Other ways to access this content:
Email to request access
Purchase this book
Related chapters
Search for more...
© Informa UK Limited, an Informa Group Company
Home | About RHO | Collections
| Free trial
| How to buy
| Promotions
| Contact us | Help & FAQs
Librarian resources
| Privacy policy
| Cookies
| Terms & conditions
