Module 1: Guidelines on PD, LGD & Defaulted Exposures
Topics
• Reduction of parameter variability
• Homogenization of PD/LGD calculation
• Data quality and representativeness
• Human judgment & margin of conservatism (MoC)
• PD estimation (development, data, drivers, philosophy)
• PD calibration (1-year DR, observed DR, LRA DR)
• LGD estimation (recoveries, collaterals, costs, calibration)
• LGD in-default & ELBE
• Application of risk parameters
• Review, documentation, impact assessment
(No exercises listed)
Module 2: Implementation of IFRS 9 in the EU
Topics
• SICR & staging
• Low credit risk exemption
• PD as proxy for lifetime
• ECL model types
• Model limitations & overlays
• ESG & climate risk influence
• IFRS 9 PD variability
• FLI incorporation (macroeconomic scenarios, non-linearity)
• Staging, measurement, overlays
• IFRS 9 LGD, IFRS 9 PD
(No exercises listed)
Module 3: Exploratory Analysis
Topics
• Data sources, target definition
• Sampling (random, stratified, rebalanced)
• EDA: histograms, QQ plot, boxplot
• Missing values & multivariate imputation
• Outliers: winsorizing, trimming, Mahalanobis
• Oversampling / SMOTE
Exercises
1. EDA Exploratory Analysis
2. Feature Engineering
3. Detection & treatment of Advanced Outliers
4. Multivariate imputation of missing values
5. Univariate analysis in percentiles (R)
6. Optimal univariate continuous-variable analysis (Excel)
Module 4: Feature Engineering
Topics
• Standardization
• Binning (equal interval, frequency, chi-square)
• WOE coding
• Variable selection
• Continuous & categorical optimization
• Information Value
(Exercises listed in Module 3)
Module 5: Machine Learning
Topics
• K-Means
• PCA & visualization
• Ensemble Learning: Bagging, Random Forest, Boosting, AdaBoost, Gradient Boosting
Exercises
1. Segmentation using K-Means
2. Credit Scoring SVM
3. Credit Scoring Boosting
4. Credit Scoring Bagging
5. Credit Scoring Random Forest (R & Python)
6. Credit Scoring Gradient Boosting Trees
Module 6: Deep Learning
Topics
• Feedforward NN, MLP
• Activation functions
• Backpropagation, gradients, Jacobians
• CNNs (filters, pooling, FC layers)
• CNN for credit scoring
• RNN, LSTM
• GANs for credit scoring
Exercises
1. Deep Learning Feedforward Credit Scoring
2. CNN for Credit Scoring
3. LSTM Credit Scoring
4. GANs Credit Scoring
Module 7: Hyperparameter Tuning
Topics
• Grid, random & Bayesian search
• Learning rate, optimization algorithms
• Activation functions
• Loss & cost functions
• Hidden layers, units, epochs
• Dropout, batch size
• SHAP interpretation
Exercises
1. Tuning XGBoost, Random Forest & SVM
2. Tuning Deep Learning models
Module 8: Scorecard Development
Topics
• Scorecard design (binary, continuous)
• WOE classification
• Reject inference
• Cut-off optimization (ROC)
Exercises
1. Scorecard creation in Excel, R, Python
Module 9: Quantum Computing & Algorithms
Topics
• Qubits, circuits, measurement
• Quantum operations & entanglement
• Deutsch algorithm
• QFT
• Hybrid quantum-classical algorithms
• Quantum ML
Exercises
1. Quantum operations
Module 10: Quantum Machine Learning for Credit Scoring
Topics
• QSVM
• VQC
• Quantum NN, QGAN
• Quantum ML for PD & scoring
Exercises
1. Quantum SVM Credit Scoring
2. Quantum FFNN for PD
3. Quantum CNN for PD
Module 11: Tensor Networks
Topics
• Entanglement
• Tensor networks for ML
• Applications in NN & SVM
• Credit scoring using tensors
Exercises
1. Credit Scoring & PD with Tensor Networks
Module 12: Probabilistic Machine Learning
Topics
• Bayesian models
• Gaussian processes
• HMM
• MCMC
• Bayesian NN
Exercises
1. Gaussian Process Regression
2. Bayesian Neural Networks
Module 14: Advanced Validation of AI Models
Topics
• Explainability, SHAP
• Global & local explanations
• Model diagnosis (accuracy, robustness, fairness)
• Regression & classification comparison
Exercises
1. Validation & Diagnosis of Advanced Scoring Models
Module 15: IRB PD – Probability of Default
Topics
• PD estimation
• Data requirements
• PD calibration
• Technical defaults
• PD PIT & TTC
• ML & DL PD
• Margin of conservatism
Exercises
1. Modeling Margin of Conservatism (PD)
Module 16: Econometric & AI PD Models
Topics
• Logistic, Probit, Cox, Log-Log
• Panel Models
• Bayesian logistic
• ML survival models
• Random Survival Forest
• Cox-XGBoost
• Deep Learning Survival
• Quantum NN
Exercises
1. Cox Regression PD
2. Panel Logistic PD
3. Bayesian Logistic Regression
4. LASSO PD
5. Random Forest Survival
6. Cox-XGBoost
7. Deep Learning Survival
8. Feedforward NN
9. Quantum NN PD
Module 17: PD Calibration IRB
Topics
• Central tendency
• Bayesian calibration
• Scaled calibration
• Smoothing PD curves
• QMM, Beta, Laplace, isotonic regression
• Gaussian Process Regression
Exercises
1. Platt Scaling & Isotonic Regression
2. Gaussian Process Regression Calibration
3. Asymmetric Laplace Calibration
Module 18: Bayesian PD & Gaussian Processes
Topics
• Bayesian PD estimation
• MCMC
• Credibility intervals
• Gaussian process regression
Exercises
1. Bayesian Logistic PD (Python)
2. PD using MCMC (R)
3. Gaussian Process Regression PD
Module 19: Low-Default Portfolio PD
Topics
• PD in LDP
• Bayesian neutral & conservative PD
• GAN synthetic data for LDP
Exercises
1. LDP PD – Confidence Interval
2. Multi-period CI PD
3. Neutral Bayesian PD
4. Conservative Bayesian PD
5. Synthetic Data for PD via GAN
Module 20: IFRS 9 PD Forecasting
Topics
• PIT PD
• TTC PD
• ARIMA, VAR, VARMAX
• LSTM, Bayesian LSTM
• Quantum LSTM
Exercises
1. PD Forecasting with VARMAX
2. PD Forecasting with Quantum LSTM
Module 21: Lifetime PD
Topics
• Vintage & EMV models
• Regression & survival models
• Markov models
• ML & DL
Exercises
1. Multinomial Regression – Lifetime PD
2. Multi-State Markov – Lifetime PD
3. Matrix ASRF – Lifetime PD
4. Quantum SVM – Lifetime PD
5. SVM – Lifetime PD
6. Deep Learning Lifetime PD
7. Quantum Deep Learning Lifetime PD
Module 22: IRB LGD (Retail)
(No exercises)
Module 23: Econometric & ML Models for LGD
Exercises
1. Econometric LGD Models
2. ML & Deep Learning LGD
3. Model Comparison Tests
Module 24: LGD IFRS 9
Exercises
1. RF LGD IFRS 9
2. Beta Regression + NN LGD IFRS 9
3. XGBoost LGD IFRS 9
4. SVM LGD IFRS 9
5. Quantum SVM LGD IFRS 9
6. Traditional NN LGD
7. Bayesian NN LGD
Module 25: IRB EAD & CCF
Exercises
1. NN CCF Model
2. SVR CCF (Python)
3. NN + Beta Regression (R)
4. Quantum SVM CCF
Module 26: IFRS 9 EAD – Contractual Options
Exercises
1. Prepayment Model (RF, R & Excel)
Module 27: Lifetime EAD for Credit Lines
Exercises
1. NN Model for Lines of Credit
2. Lifetime EAD Model
Module 28: PD Backtesting
Exercises
1. PD Backtesting in Excel
Module 29: LGD Backtesting
Exercises
1. LGD Model Performance Comparison
Module 30: EAD Backtesting
Exercises
1. EAD Model Performance Comparison
Module 31: Deep Learning for Macroeconomic Dynamics
Exercises
1. Neural Networks for Bellman Equation
Module 32: Deep Learning for Macroeconomic Projections
Exercises
1. Charge-off Model with VAR & VEC
2. PD Forecasting with Bayesian LSTM
3. Pandemic Forecasting with RNN LSTM
4. PD Forecasting with Quantum NN
Module 33: Stress Testing PD & LGD
Exercises
1. Credit Portfolio View – PD Stress Test
2. Bayesian LSTM PD Stress Test
3. VQR Quantum Regression PD Stress Test
4. MARS Stress Test PD
5. LASSO Stress Test
Module 34: Corporate Portfolio Stress Testing
Exercises
1. Corporate PD Stress Test (ASRF + Transition Matrix)
Module 35: ECL IFRS 9 Stress Testing
Exercises
1. Internal ECL Stress Test (R & Excel)
Module 36: Quantum Stress Testing
Exercises
1. DGSE Model with Deep Learning
2. Quantum vs Classical Monte Carlo
Module 37: Economic Capital Models
Exercises
1. Portfolio EL/UL/Correlation/EC
2. CreditRisk+ in SAS
3. CreditMetrics in Excel & R
4. Single-Factor EC Model
Module 38: Quantum EC Models
Exercises
1. Quantum EL, VaR & ES
Module 39: Climate Risk in Credit Risk
Exercises
1. PD & DD for Transition Risk
2. PD & DD for Physical Risk
.png)