Course Objective & Scope
Standard predictive models learn historical correlations, yet intervention questions such as "how much will sales rise if we cut price?" or "how will raising this credit limit affect default?" cannot be answered from correlation alone. This programme covers causal graphs (DAGs), confounders, selection bias, propensity score matching and weighting (IPW), Double/Debiased Machine Learning (DML) and heterogeneous treatment effects (CATE) with Causal Forests.
When randomised trials are impossible, Difference-in-Differences, Synthetic Control and Instrumental Variables are applied hands-on using DoWhy and EconML. Participants produce a treatment-effect and refutation report on observational enterprise data.
Who Should Attend?
- Senior data scientists and economists/econometricians
- Pricing, marketing effectiveness, risk policy and product analytics teams
- Strategy and decision-science specialists
- Prerequisite: statistics, regression and Python machine learning experience
Key Takeaways
- Map a business question as a causal DAG and separate confounders from colliders and mediators
- Estimate Average Treatment Effects (ATE) in observational data using IPW, Doubly Robust and Double ML
- Estimate individual and segment-level treatment effects (CATE) with EconML and Causal Forests
- Evaluate policy or rollout impacts using Difference-in-Differences and Synthetic Control
- Stress-test causal estimates with DoWhy refutation checks and unobserved-confounder sensitivity
Curriculum
Causal Framing, Potential Outcomes and DAGs
- Prediction vs intervention, Simpson's paradox and selection bias
- The backdoor criterion, confounders, mediators and colliders
Estimating ATE from Observational Data
- Propensity scores, IPW and Doubly Robust (AIPW) estimators
- Checking positivity/overlap and weight trimming
Double ML and Heterogeneous Effects (CATE)
- Double/Debiased ML and meta-learners (S/T/X/R-learner) with EconML
- Causal Forests for personalised policy and price elasticity
Quasi-Experimental Designs
- Difference-in-Differences (DiD) and parallel trends diagnostics
- Synthetic Control, Regression Discontinuity and Instrumental Variables
Validation, Refutation and Policy Learning
- Placebo treatments, random common causes and subset tests in DoWhy
- Learning optimal targeting policies under budget constraints
Other Courses in This Category
-
Collaborative model training across institutions or edge devices without centralising raw data: Flower framework, FedAvg/FedProx, non-IID data, differential privacy (Opacus) and secure aggregation.
- Live Course
- 5 modules
-
Intro to Machine Learning - Regression Models
Hands-on workshop covering Linear, Ridge, Lasso, and Polynomial regression models, feature engineering, MSE/RMSE/MAE/R² evaluation metrics, and regularization techniques.
- Live Course
- 3 modules
-
Incremental learning on streaming data with River, prequential evaluation, concept drift detection (ADWIN, Page-Hinkley) and contextual bandits for continuous adaptation in real-time systems.
- Live Course
- 5 modules
-
Modelling relational data with NetworkX and PyTorch Geometric: centrality and community algorithms, graph features, Node2Vec and GCN/GraphSAGE/GAT for node, link and subgraph prediction in fraud rings and supply networks.
- Live Course
- 5 modules
-
In-context learning on tabular data with TabPFN v2, plus zero-shot and fine-tuned time-series forecasting with Chronos, TimesFM, Moirai and Lag-Llama, benchmarked against GBDTs and classical models.
- Live Course
- 5 modules
-
Supervised and unsupervised learning, train/test splits, overfitting, core metrics and first models in scikit-learn, so technical teams can frame a business problem as an ML task and build a sound baseline.
- Live Course
- 5 modules
Get in touch
