Page under construction.
The course covers advanced statistical methods for machine learning. Module 1 introduces advanced statistical foundations for machine learning, focusing on robustness, decision-making under uncertainty, calibration, Bayesian and causal inference, and modern representation techniques. Module 2 addresses high dimensional and large-scale data challenges, emphasizing feature selection, dimension reduction, and strategies for managing massive or imbalanced datasets. Theoretical concepts are complemented by practical exercises and project work using the R or Python programming languages.
Module 1 (INF/01, 3 ECTS)
– Noise and robust statistics
– Statistical decision theory
– Classifier calibration, conformal methods and learning to reject/defer
– Bayesian inference
– Causal inference: structured causal model, potential outcome model
– PCA and embeddings in LLM.
Module 2 (SECS-S/01, 3 ECTS)
– Feature selection and regularization techniques for high-dimensional Linear and Generalized Linear Models
– Feature screening algorithms for ultra-high dimensional supervised problems
– Supervised dimension reduction; Sufficient Dimension Reduction and related techniques
– Subsampling/partitioning approaches for ultra-high sample sizes
– Under- and oversampling approaches for data rebalancing.