Big Network Data
Big Network Data
Syllabus:
A course on network science driven by data analysis. The class will focus on both theoretical and empirical analysis performed on real data. Basic principles underlying the structure and dynamics of complex networks including technological networks, social networks, information networks, biological networks, economic networks, financial networks. Fundamentals of network theory, network representations, elements of graph theory. Metrics, degree distributions, centrality, clustering, connectivity, diameter, components, modularity. Generative models, Erdos-Renyi graphs, power law graphs, configuration model, percolation, small worlds, random connection models. Fitting distributions from data, unbiased sampling, data fitting vs. model validation, the scale-free debate. Processes on networks, consensus, group formation, voting, epidemics, segregation, community detection. Extracting information from networks, learning, feature discovery, dimensionality reduction. Students will be exposed to a number of state of the art software libraries for network data analysis and visualization via the Python notebook environment.
Prerequisites:
Familiarity with Python programming. Basic probability and statistics.
Books:
Lectures will draw material from the following books. We do not strictly follow one, so you will have to be up to date with what is discussed in class. Additional reading material will be posted on-line in the handout section.
•Networks. by M.E.J. Newman. Oxford.
Graduate level book focusing on the theory of complex network.
•Networks Corwds and Markets. By D. Easley and J. Kleinberg. Cambridge.
Undergraduate level book giving a very good, broad introduction.
•A First Course In Network Theory. By E. Estrada, P.A. Knightung Chiang. Oxford.
Somewhere in between, an introductory book with focus on teaching fundamentals.
Class times and location (Fall 2017):
TuTh 3:30p-4:50p WLH 2207
TA: Amey Paranjape ajparanj@eng.ucsd.edu
Office hours: THURSDAY 10:30 am - 1:30 pm Jacobs 5101E
Prof. Franceschetti Office Hours: ATK4302 Immediately after class or by appointment
Class Organization and Grading:
There will regular homework assignments including both programming for data analysis and pen and paper exercises. Class grade is based on your performance on these assignments (70%) and a final poster session at the end of the class (30%) where you will present and expand upon one of the topics in the homeworks.
Image drawing by prof. R. Graham