Παρουσίαση/Προβολή

Supervised and unsupervised machine learning (ML) methods. Linear and non-linear ML models and their statistical and information theoretic foundations. Parameter estimation with classical maximum likelihood and Bayesian methods. Regularization, model selection. Data clustering (hierarchical, k-means,  etc.). Neural networks (supervised and unsupervised). Bayesian networks, graphical models, Hidden Markov models. Generative models. Parallel processing for high-performance machine learning. Privacy preservation, fairness, and bias in biomedical AI. Distributed federated learning methods. Applications in computational biology, bioinformatics, biotechnology, and healthcare, such as: decoding  DNA sequences, analysis of gene expression profiles (single cell RNAseq data analysis), processing of proteomic and metabolomic data. Gene and protein interaction networks. Deriving system models from multidimensional heterogeneous data, model comparison, multidimensional data visualization, applications in systems biology. Modeling the dynamics of biological systems, mechanistic models, such as ligand-receptor interaction or the spread of  Covid-19. Stochastic dynamics, stochastic simulation. The synergistic role of data-driven (static) and theory-driven (dynamic) models for systems biology. Towards digital twins for healthcare. Applications of systems biology in drug design. Development of computational ML pipelines to analyze real-world data using Python and R programming. Term project and class presentation required.