Course Abstract

Training duration: 4 hr (Hands-on)

Networks, also known as graphs are one of the most crucial data structures in our increasingly intertwined world. Social friendship networks, the world-wide-web, financial systems, infrastructure (power grid, streets), etc. are all network structures. Knowing how to analyze the underlying network topology of interconnected systems can provide an invaluable skill in anyone's toolbox. This tutorial will provide a hands-on guide on how to approach a network analysis project from scratch and end-to-end: how to generate, manipulate, analyze and visualize graph structures that will help you gain insight about relationships between elements in your data.

DIFFICULTY LEVEL: BEGINNER

Learning Objectives

  • Understand the basics of graphs/networks properties and analysis, including what can you use it for and how

  • Learn how to generate basic network types, and the most often encountered network models in real data. Next, discover the most informative network measures to understand network structures and behaviors

  • Extract and interpret information about real public social network data by building, analyzing and visualizing it to gain understanding about its structure and behaviors

Instructor Bio:

Matt Brems

Senior Inventive Scientist | AT&T Chief Data Office

Noemi Derzsy, PhD

Noemi Derzsy is a Senior Inventive Scientist at AT&T Chief Data Office within the Data Science and AI Research organization. Her research is centered on understanding and modeling customer behavior and experience through large-scale consumer and network data, using machine learning, network analysis/modeling, spatio-temporal mining, text mining and natural language processing techniques. Prior to joining AT&T, Noemi was a Data Science Fellow at Insight Data Science NYC and a postdoctoral research associate at Social Cognitive Networks Academic Research Center at Rensselaer Polytechnic Institute. She holds a PhD in Physics, MS in Computational Physics, and has a research background in Network Science and Computer Science. Noemi is also involved in volunteering in the data science community. She is a NASA Datanaut and former organizer of the Data Umbrella meetup group and NYC Women in Machine Learning and Data Science meetup group.

Course Outline

Module 1: The Unreasonable Effectiveness of Deep Learning 


●        Training Overview 

●        A Brief History from Graph Theory to Network Science

●        Real-World Applications of Networks/Graphs Overview

●        Basic Network Structural Properties

●        Graphs in Python with NetworkX


Module 2: Generate & manipulate graph structures


●        Create, modify and delete graphs 

●        Node, edge properties and structure

●        Create graph structure from datafile 

●        Weighted graphs

●        Directed graphs

●        Multigraphs

●        Bipartite graphs



Module 3: Analyze and visualize networks 


●        Structural properties analysis

●        Node degree, average degree, degree distribution

●        Clustering, coefficient, triangles

●        Centrality measures

●        Components

●        Assortativity

●        Network visualization with NetworkX

●        Network visualization with nxviz

●        Visualize subgraphs

●        Network visualization with node attributes

Background knowledge

  • Basic Python

  • Jupyter Notebooks

  • Installation of NetworkX package

This course could be useful for:

  • Social media platforms

    Network analysis and modeling is used by online social media companies (i.e. Facebook, Twitter) to study opinion formation and influencing in social networks. Graph-based methods are also used to suggest new contacts on the platform or to recommend new products to customers, based on the products their online friends are interested in

  • Finance

    Financial systems, the interconnected stock market can be modeled and analyzed using network/graph methods.

  • Infrastructure

    Graph-based analysis and modeling are crucial in solving transportation system optimization problems, such as optimizing power-grid systems, airline or ground traffic flow and determine shortest paths, the most cost-efficient routes between destinations (i.e. Google Maps).

  • Healthcare

    Contact tracing skills and the ability to analyze and model infectious disease spreads, are all essential applications of networks/graph, especially during the COVID-19 pandemic.