Credit Risk Scoring & Decision Making by Global Experts

Welcome to our Credit Risk Modeling course! By the end of this course, you will have a solid understanding of credit risk models and their applications in the industry. This video will provide you with a clear outline of the course structure, helping you navigate through each module and lesson. Get ready to enhance your skills and apply them to real-world scenarios!

In this video, "Setting Up Your Computer," you will learn how to install Anaconda, a powerful open-source distribution of Python and R for scientific computing. Anaconda simplifies package management and deployment, providing you with all the tools you need for data science, machine learning, and credit risk modeling.

In this video, "Overview of Credit Risk Models," we will introduce the three core components of credit risk modeling: Probability of Default (PD), Loss Given Default (LGD), and Exposure at Default (EAD). These models form the foundation for assessing and managing credit risk in financial institutions.

In the "Applications in the Industry", we will explore how credit risk models are utilized across various sectors.

In the "Python Codes" section, you will find all the essential course code and project solutions that will be used throughout the course. This section serves as a comprehensive resource, providing you with practical examples and detailed code implementations that align with the concepts we will cover. Whether you're looking to follow along with the lessons or seek solutions to the projects, this section will be your go-to reference for hands-on learning and application.

In the "Documents" section, you will find important materials that are crucial for understanding credit risk. These documents include detailed explanations, theoretical concepts, and industry standards related to credit risk modeling.

In the "Introduction to PD Models" section, we will delve into Probability of Default (PD) models, a fundamental component of credit risk assessment. Here, you will learn about the basic concepts, the importance of PD models in predicting the likelihood of default, and how they are applied across various financial contexts. This section will provide a solid foundation for understanding how PD models are built, validated, and used to manage and mitigate credit risk effectively.

In the "Example Case" section, we will explore a real-world scenario to demonstrate how credit risk models are applied.

In the "Application vs. Behavioral Scorecards" section, we will delve into the differences between application scorecards and behavioral scorecards. Application scorecards are used at the initial stage of evaluating a new customer's creditworthiness, based on information provided during the application process. In contrast, behavioral scorecards are used to assess the credit risk of existing customers, focusing on their historical payment behavior and account management. This section will help you understand when and how to use each type of scorecard effectively in credit risk assessment.

In this video, we will provide a detailed description of the dataset used in this course.

In this video, we'll demonstrate how to load data into the Python environment using Pandas.

In this video, we’ll explore the concept of data quality, focusing on its key aspects such as accuracy, completeness, consistency, and reliability. We'll discuss methods for assessing and ensuring high data quality to support effective analysis and decision-making.

In this video, we'll cover data cleaning techniques, including methods for handling missing values, correcting errors, and standardizing data. You'll learn how to prepare your dataset for analysis by ensuring it is accurate, complete, and consistent.

In this video, we'll dive into Exploratory Data Analysis (EDA), covering techniques for summarizing and visualizing your data. We’ll explore methods to uncover patterns, identify anomalies, and gain insights to guide further analysis

In this video, we'll focus on Exploratory Data Analysis (EDA) based on time. We’ll explore methods for analyzing time series data, identifying trends, seasonal patterns, and anomalies to better understand how data evolves over time.

In this video, we'll review sector best practices for data preprocessing. We’ll cover effective techniques for data cleaning, transformation, and integration, ensuring high-quality data that supports robust analysis and decision-making.

In this video, we'll cover data transformation techniques, including methods for scaling, encoding, and aggregating data. You’ll learn how to prepare and modify your data to improve its usability and effectiveness for analysis.

In this video, we'll apply data transformation techniques in practice. We’ll walk through real-world examples of scaling, encoding, and aggregating data to enhance its quality and prepare it effectively for analysis.

In this video, we'll explore sector best practices for data transformation. We’ll discuss industry-specific techniques to ensure high-quality, actionable insights.

In this video, we'll review various data splitting methods, including train-test and train-validation-test splits. We’ll discuss their importance for model training and evaluation, and how to choose the right approach for your analysis.

In this video, we’ll demonstrate data splitting in practice, showcasing how to implement train-test and train-validation-test splits. We’ll provide step-by-step examples and tips to ensure effective model training and evaluation.

In this video, we’ll provide an overview of feature selection methods and explore sector best practices. We’ll discuss various techniques for identifying the most relevant features and ensure effective model performance through industry-standard approaches.

In this video, we'll explore correlation elimination techniques, focusing on methods to identify and remove highly correlated features. This process helps to reduce redundancy and improve the performance and interpretability of your models.

In this video, we’ll apply correlation elimination techniques in practice. We’ll demonstrate how to identify and remove highly correlated features from a dataset, enhancing model efficiency and reducing redundancy.

In this video, we'll cover the concept of Information Value (IV), including how to calculate and interpret it. We’ll explore its role in assessing the predictive power of features and its application in feature selection for modeling.

In this video, we’ll demonstrate the practical application of Information Value (IV). We’ll show how to calculate IV for features, interpret the results, and use this information to make informed decisions in feature selection.

In this video, we’ll explain the Univariate Gini coefficient, covering its calculation and interpretation. We’ll discuss how it measures the discriminatory power of a single feature and its role in evaluating feature importance for predictive modeling.

In this video, we’ll apply the Univariate Gini coefficient in practice. We’ll walk through the calculation and interpretation of Gini scores for individual features and demonstrate how to use this information to assess feature effectiveness in a dataset.

In this video, we’ll introduce survival analysis, focusing on techniques for analyzing time-to-event data. We’ll cover key concepts such as survival functions, hazard rates, and how to interpret and apply these methods to assess and predict event outcomes.

In this video, we’ll apply survival analysis techniques in practice. We’ll demonstrate how to analyze time-to-event data, calculate survival functions and hazard rates, and interpret the results to make informed predictions and decisions.

In this video, we’ll cover logistic regression, including its fundamentals, how it models binary outcomes, and how to interpret coefficients. We’ll also demonstrate how to implement logistic regression and evaluate its performance.

In this video, we’ll apply logistic regression in practice. We’ll walk through the implementation process, including model fitting, evaluating performance, and interpreting the results to make predictions on binary outcomes.

In this video, we’ll explore methods for explaining logistic regression models. We’ll cover techniques such as feature weights to understand how the model makes predictions and the influence of individual features.

In this video, we’ll walk through the code for calculating variable weights in a model. We’ll cover how to extract and interpret feature weights, providing insights into the relative importance of each variable in the model’s predictions.

In this video, we’ll delve into understanding model coefficients. We’ll explain how to interpret coefficients, their impact on predictions, and their role in assessing feature importance in various types of models.

In this video, we’ll focus on logistic regression with a focus on maximizing the Gini coefficient. We’ll explore how to optimize the model for the highest discriminatory power and evaluate its performance using the Gini metric.

In this video, we’ll demonstrate how to make predictions using a logistic regression model optimized for the maximum Gini coefficient. We’ll show how to apply the model to new data and interpret the results to assess its predictive performance.

In this video, we’ll explain K Fold Cross Validation, including its process and benefits. We’ll demonstrate how to split data into K subsets, train and evaluate models on different folds, and use this technique to ensure robust and reliable model performance.

In this video, we’ll apply K Fold Cross Validation in practice. We’ll demonstrate how to implement this technique, train models on different folds, and evaluate performance to ensure the model's robustness and generalizability.

In this video, we’ll explore sector best practices for classical Probability of Default (PD) models. We’ll cover effective techniques for model development

In this video, we’ll provide an overview of feature selection techniques for advanced data science.

In this video, we’ll explore feature selection using Random Forest. We’ll demonstrate how to use feature importance scores from Random Forest models to identify and select the most relevant features, improving model accuracy and efficiency.

In this video, we’ll cover feature selection using Shapley values. We’ll explain how to calculate Shapley values to determine the contribution of each feature, and how to use this information to select the most impactful features for your model.

In this video, we’ll explore Permutation Importance for feature evaluation. We’ll demonstrate how to calculate and interpret feature importance scores by measuring the impact of feature shuffling on model performance, helping to identify key predictors.

In this video, we’ll provide an overview of XGBoost, including its key features and advantages. We’ll cover its boosting algorithm, model performance benefits, and how to implement XGBoost for improved predictive accuracy and efficiency.

In this video, we’ll walk through implementing XGBoost from scratch. We’ll cover the core concepts of gradient boosting, the construction of decision trees, and how to code XGBoost algorithms step-by-step to build your own boosting model.

In this video, we’ll explore XGBoost hyperparameter tuning. We’ll cover techniques for optimizing parameters like learning rate, maximum depth, and subsample ratio to enhance model performance and achieve the best results.

In this video, we’ll provide an overview of neural networks. We’ll cover their basic structure, key components such as layers and activation functions

In this video, we’ll build a neural network from scratch. We’ll walk through the process of coding the network architecture, implementing forward and backward propagation, and training the model to solve a credit risk, providing a foundational understanding of neural networks.

In this video, we’ll cover neural network hyperparameter tuning. We’ll explore techniques for optimizing parameters such as learning rate, batch size, and number of layers to enhance model performance and achieve better results.

In this video, we’ll delve into model ensembling techniques. We’ll explain methods such as bagging, boosting, and stacking to combine multiple models, enhancing overall performance and robustness in predictive tasks.

In this video, we’ll demonstrate code for implementing ensemble methods. We’ll walk through examples of combining models using techniques like bagging, boosting, and stacking, showing how to improve predictive accuracy and model performance.

In this video, we’ll explore advanced data science techniques for credit risk management. We’ll cover methods such as advanced feature selection, machine learning models, and model validation strategies tailored to enhance credit risk assessment and decision-making.

In this video, we’ll outline model selection methodology, covering techniques for choosing the best model based on performance metrics, validation strategies, and practical considerations. We’ll demonstrate how to evaluate and compare different models to ensure optimal results.

In this video, we’ll walk through the code for model selection. We’ll demonstrate how to implement and compare various models using performance metrics, validation techniques, and selection criteria to identify the best model for your specific needs.

In this video, we’ll cover the development of a rating scale. We’ll explore the process of creating and implementing a scale for evaluating performance or risk, including defining criteria, assigning ratings, and ensuring consistency and accuracy in assessments.

In this video, we’ll demonstrate code for implementing a rating scale. We’ll walk through the process of coding the rating logic, applying it to data, and ensuring accurate and consistent ratings based on predefined criteria.

In this video, we’ll cover the process of score generation. We’ll demonstrate how to calculate and generate scores based on data inputs, including techniques for scoring models, interpreting results, and applying scores to evaluate performance or risk.

In this video, we’ll explore best practices for developing a rating scale within a sector context. We’ll cover key steps in creating a robust scale, including defining criteria, assigning ratings, and ensuring consistency and relevance for effective assessments.

In this video, we’ll dive into model calibration. We’ll explain techniques for adjusting model predictions to better reflect true probabilities, including methods for calibration curves and adjustments to improve model accuracy and reliability.

In this video, we’ll explore Bayesian calibration. We’ll cover how to apply Bayesian methods to adjust and refine model predictions, incorporating prior knowledge and updating estimates to enhance accuracy and reliability.

In this video, we’ll discuss regression calibration. We’ll cover techniques for adjusting logistic regression model predictions to improve their accuracy, including methods for recalibrating outputs and aligning predictions with observed data.

In this video, we’ll review sector best practices for model calibration.

In this video, we’ll cover the basics of model validation and explore sector best practices. We’ll discuss fundamental concepts for assessing model performance, including validation techniques and industry-specific standards to ensure reliable and accurate results.

In this video, we’ll explore validation metrics.

In this video, we’ll dive into ROC AUC, explaining the Receiver Operating Characteristic curve and the Area Under the Curve metric.

In this video, we’ll dive into Gini.

In this video, we’ll dive into Kolmogorov-Smirnov.

In this video, we’ll dive into Confusion Matrix.

In this video, we’ll dive into PSI & SSI

In this video, we’ll dive into Variance Inflation Factor.

In this video, we’ll dive into HHI - Adjusted HHI.

In this video, we’ll dive into Anchor Point.

In this video, we’ll dive into Chi-Square Test.

In this video, we’ll dive into Binomial Test.

In this video, we’ll dive into Adjusted Binomial Test.

In this video, we’ll dive into Model Validation Thresholds.

In this video, we’ll explore a case study of a U.S.-based financing company. We’ll analyze their approach to financial modeling, risk assessment, and decision-making, highlighting key strategies and insights gained from their real-world applications.

In this video, we’ll explore a case study of a UK -based fintech startup. We’ll analyze their approach to financial modeling, risk assessment, and decision-making, highlighting key strategies and insights gained from their real-world applications.

In this video, we’ll guide you through a final project using real-world data.

The "Final Assessment" section evaluates your comprehensive understanding of credit risk scoring concepts through a detailed test or project to confirm your mastery of the material.