Advanced credit risk analytics using Python


Instructor led online course 

Learn more  Enrol now 

Join any of our batches

Weekday batch (Click here to see the dates)
Weekend batch (Click here to see the dates)

15 hours of intensive training on advanced credit risk analytics

Pay in INR or USD as per your choice

₹20,000

Pay Now 

$300

Pay Now

Apply practical methods and build your own models using Python

Statistical techniques

Learn how statistical techniques are practically applied in credit risk analytics

Modelling techniques

Get up to speed with the latest modelling techniques and practices

Build on Python

Learn to build your own automated models

Stay updated

Keep step with the evolving credit risk practices

Who is this course meant for?

Credit risk practitioners
Professionals in the Business Intelligence Units of banks
Credit appraisal teams
Internal audit/  governance teams involved in model risk management
Model risk management and model validation units

Pre-requisites for the course

Basics of Python and setup of Python environment
  • Understanding of the basics of Python
  • Installed Python environment
  • IDE to be used for Python installed
  • Understanding of basic Python functions and syntax
  • Understanding of data types in Python
  • Leveraging Python packages (Pandas dataframes, Numpy and statistical analysis)
Statistical knowledge
  • Understanding of basic quantitative measures (mean, median, standard deviation, statistical hypothesis)
  • Foundational knowledge on statistical modelling techniques such as moving average, trend analysis
  • Random variables and probability distributions
  • Apply concepts of statistical analysis through analysis of sample and population data characteristics
  • Understanding use of correlation (Pearson coefficient) to identify data relationships
  • Linear and optimisation models (growth and decay processes, discrete and continuous)
  • Probabilistic models (Markov chains, tree-based models)

Training coverage

Coverage aspects:
− Requirement of credit risk models
− Different use cases and outputs
− Different types of credit risk models

Learning outcome:
− Understanding of types of modelling requirements for credit risk assessment
− Understanding of use cases 

Coverage aspects:
− Defining model use
− Model governance setup
− Model development
− Model validation (statistical, performance, and qualitative methods)
− Model health and risk monitoring

Learning outcome:
− Overview of modelling lifecycle, model efficacy testing, and governance needed for model maintenance

Coverage aspects:
− Selecting the sample
− Types of variables and description specific to retail credit portfolios
− Missing values (imputation schemes)
− Outlier detection and treatment (box plots, z-scores, truncation, etc.)
− Exploratory data analysis
− Categorization (chi-squared analysis, odds plots, etc.)
− Weight of evidence (WOE) coding and information value (IV)
− Segmentation
− Reject inference (hard cut-off augmentation, parcelling, etc.)

Learning outcome:
− Understanding of data checks and data preparation steps (using Python tools) required for performing modelling analysis using retail portfolio data

Coverage aspects:
− Description of portfolio
− Description and application of models
− Types of scorecards (application, behavioural and dynamic)
− Feature engineering (Developing target variable from combination of existing variables)
− Modelling techniques available
− Impact / Risk of using scoring models
− Pricing and profitability computation based on scores (Risk based profitability)

Learning outcome:
− Understanding of retail portfolio and techniques available for developing scoring models
− Hands on exercise in Python along with retail scoring models

Coverage aspects:
− Regression analysis - type and usage
− Data preparation for regression analysis
− Interpreting results
− Correlation analysis
− Linear and logistic regression analysis

Learning outcome:
− Understanding of advanced statistical concepts
− Hands on Python work along examples for advanced statistical concepts used for retail scoring and other credit risk analysis 

Coverage aspects:
− Perform regression analysis using Python

Learning outcome:
− Hands on case study for using regression technique for actual business use case

Coverage aspects:
− Development of regression-based retail scoring model
− Data cleaning and transformation to required input structure for regression
− Techniques for variable selection (significance testing, backward / forward iterations)
− Techniques for model optimization

Learning outcome:
− Understanding of scorecard development process for a retail portfolio using Python

Coverage aspects:
− Assignment related to data analysis and regression analysis using Python - Explanation and provide sample data

Learning outcome:
− Explanation of day end assignment provided for participants to apply concepts

Coverage aspects:
− Statistical testing of regression models (linearity, multicollinearity, autocorrelation, stationarity, heteroskedasticity)
− Performance testing of scoring models (Adj R2, MAPE, RMSE)
− Monitoring (back-testing) model performance
− In-sample / out-of-sample / out-of-time testing for model predictability (keeping in mind structural breaks)

Learning outcome:
− Understanding of validation of an application scorecard for a retail portfolio process using Python

Coverage aspects:
− Introduction to Machine Learning models
− How machine learning models compare with traditional analysis
− Types of Machine Learning models (Supervised Learning, Unsupervised Learning, Reinforcement Learning)
− Use cases where ML models are used

Learning outcome:
− Overview of ML models
− Understanding of types of ML models
− Understanding of use cases where ML models are often used 

Coverage aspects:
− Introduction to Decision Tree and Random Forest
− Hyperparameter tuning
− Common algorithms used to optimize hyperparameters
− Model fit assessment and finetuning
− Case study: Classification modelling using Decision Tree and Random Forest

Learning outcome:
− Understanding of types of Decision Tree and Random Forest models
− Understanding of hyperparameter tuning process
− Hands on work along exercise for implementing Decision tree and Random Forest models 

Coverage aspects:
− Introduction to SVM
− SVM model implementation using Python
− Model fit assessment and finetuning
− Case study: Classification modelling using SVM model

Learning outcome:
− Understanding of SVM algorithm
− Hands on work along exercise for implementing SVM algorithm 

Coverage aspects:
− Introduction to Boosting techniques and benefits derived
− XGBoost algorithm introduction and implementation
− Model fit assessment and finetuning
− Case study: Classification modelling using XGBoost

Learning outcome:
− Overview of Boosting techniques
− Understanding of XGBoost algorithm
− Hands on work along exercise for implementing XGBoost algorithm 

Coverage aspects:
− Introduction to unsupervised learning techniques
− Principle Component Analysis
− Clustering analysis
− Neural Networks
− Case study: Using unsupervised learning techniques for customer pooling

Learning outcome:
- Conceptual understanding of unsupervised learning techniques
− Use of unsupervised learning models

Coverage aspects:
− Assignment related to ML modelling analysis using Python - Explanation and provide sample data

Learning outcome:
− Explanation of day end assignment provided for participants to apply concepts

Total training hours

15 hours across 3 days 

Assessment involved

Yes - at the end of the course

Certification

On completion of assessment

Register yourself today!

 Enrol now 
Company
Businesses
Info Center


Copyright © 2024 ACIES

ACIES, Acies Consulting, Acies TechWorks, Acies Ventures, Acies LightHouse, Kepler, Kore, Callisto, Carpo, Revolutio and Antares are registered trademarks of Acies LLP, Acies Consulting Pte Ltd. and its subsidiaries in different countries and territories. Unauthorized use, misrepresentation or misuse of trademarks is strictly prohibited and subject to legal action under various laws.