911±¬ÁÏÍø

TheÌýApplied Data ScienceÌýandÌýVisualizationÌýforÌýComplex SystemsÌýmoduleÌýprovides an interdisciplinary, mixed methods approach toÌýdata analysis, visualization, andÌýhypothesis generationÌýfor understandingÌýcomplex systems,Ìýthe strategies for governing these systems, and the data produced by theseÌýprocesses. ÌýGiven the rich and diverse data produced by these systems and the challenges of unravellingÌýcomplex relationships, this course leverages a unique combination of traditional statistical learning,Ìýdata-based storytelling, and visualization toÌýcreate the kinds of thick narratives about system behaviour that have historically requiredÌýintensive qualitative analysis of the system and actors engaged in that system.ÌýÌýThis module starts with the well-known ELT (extract, load, transform) model, highlightingÌýthe role of data integration, provenance, and governance in the analysis of modern complex systems.ÌýVisualization isÌýpresentedÌýas a form of modeling, with visualization concepts and strategies integratedÌýwith data transform, exploratory data analysis (EDA), and methods developmentÌýconcepts and skillsÌýthroughout the module. ÌýThe first four sessions of the moduleÌýfocus on data integration, transforms,Ìýmanaging data provenance, data governance, and the fundamentals of package development for reproducible research. ÌýTheÌýremain six sessionsÌýintroduce clustering (k-means, hierarchical), social network analysis (with a focus on cliques), and text mining (sentiment analysis, topic modeling,Ìýnetwork analysis ofÌýcorpi)Ìýas methods for performing EDAs for hypothesis generation.ÌýA key theme throughout the course is developing visualizationsÌýappropriate to the stage of analysis and hypothesis generation, and the audience. ÌýAt the conclusion of this module, students will have a strong grasp on the workflows, tools, and methods necessary toÌýintegrate diverse datasets,Ìýconscientiously developing and distinguishing betweenÌývisualizations and analyses that facilitateÌýEDA,ÌývalidatingÌýpatterns identified, communicatingÌýwith system experts, and, importantly for studentsÌýworking at the interface of complex systems and policy, visualizations and indicatorsÌýthat communicate essential systemÌýbehavioursÌýto non-technicalÌýactorsÌýsuch as policy makers, regulators, and c-levels, while preserving the fidelity of underlying models.