EE E6885:  Topoics in Signal Processing: Network Science

COURSE BENEFITS:

  • Students will gain knowledge and hands-on experience about analyzing network data – e.g., social, information networks. It serves as an introductory course for graduate students who are interested in the research area of network science.
  • Gain knowledge with fundamental mathematical analysis of network data. Understand foundations, such as network feature extraction, partitioning (community extraction), visualization, sampling, estimation, etc.
  • Survey latest applications, such as social influence, information diffusion, privacy, security, economy issues, behavior and cognitive understanding, etc.  
  • Gain hands-on experience in network analytics through homeworks and final project.

PROFESSOR CHING-YUNG LIN:

Ching-Yung Lin is the IBM Lead of the Social and Cognitive Network Science Research Center.  He joined IBM Watson Research Center in 2000. He is also an Adjunct Associate/Full Professor at Columbia University since 2005, and an Affiliate Assistant/Associate Professor at the University of Washington 2003-9. His research interest focuses on multimodality signal processing and understanding, data and network mining, and security. Dr. Lin is a keynote speaker of Web 2.0 Expo, New York, 2009. His research work on the “Value of Social Networks” was selected as “Top Story of the Week” of BusinessWeek Magazine, April, 2009. Dr. Lin is the Chair of IEEE CAS Society Multimedia Technical Committee, 2010-2011, the General Chair of IEEE Intl. Conf. on Multimedia and Expo (ICME), 2009, and the Steering Committee Chair of ACM SIG Health Informatics (IHI), 2010-2012. Dr. Lin is a Fellow of the IEEE.

 

APPLICABLE DEGREE PROGRAMS:

  • Recommended for MS or Ph.D. in Electrical Engineering, Computer Science or any discipline require analyzing network data.
  • Most courses 4000-level and above can be credited to all degree programs.  All courses are subject to advisor approval.

COURSE FEES:

  • None

Lecturer/Manager:

Ching-Yung Lin

 

Office Hours:

Monday 9:30 - 10:00pm or by appointment

Office Location/Phone:

Mudd 227

Email Address:

cylin {at} ee {dot} columbia {dot} edu

 

Day & Time Class 
Meets on Campus:

Monday 7:00pm - 9:30pm

Location:

Mudd 227

Credits for course:

3

Class Type:

Lecture

Prerequisites:

Basic courses in probability & statistics such as E3658 (Probability) and signal processing such as E3801 (Signals and Systems). Students need to know at least one programming language or statistical tool (Matlab, C, Java, C++, Perl, Python, SPSS, R) for finishing homeworks. Students outside the Engineering School (e.g., GSAS, Business School, Law School, etc.) are also welcome.

Description:

Network as a new scientific discipline is emerging. Entities -- people, information, societies, nations, devices -- connect to each other and form all kinds of intertwined networks. Social-informatic networks are becoming part of our daily life. How do we analyze these “Big Network Data”?

 

Researchers from multiple disciplines -- electrical engineering, computer science, sociology, public health, economy, management, politics, laws, arts, physics, math, etc – are interacting with each other to build up common grounds of network science. Network theories are being formed for describing the dynamics, behaviors, and structures. A systematic mathematical formalism that enables predictions of network behavior and network interactions is also emerging. Trans-disciplinary approaches are usually required to lay the foundations of this science and to develop the requisite tools.

 

Network data analysis will become essential for graduate or senior students pursuing advanced understanding in Signal Processing. For its interdisciplinary nature, it is also suitable for students in other disciplines conducting researches in various forms of networks.

 

Students shall learn fundamental network analysis concept, math, and tools, and gain hand-on experiences. Students will need to choose final project topic of their interest and conduct comprehensive study or preliminary research.

TA:

Xiao-Ming Wu

Office Hours:

Fridays 3-5pm

Office Location/Phone:

7LE3 Schapiro Building (CEPSR)

Email Address:

xw2223 {at} columbia {dot} edu

 

Required Text(s):

Textbook:  E. Kolaczyk, “Statistical Analysis of Network Data,” Springer, 2009

 

Reference Text(s):

class notes, and reference papers will be available at http://www.ee.columbia.edu/~cylin/course/netsci/NetSci_syllabus.html

Homework(s):

four assignments including analytical questions and programming

Project(s):

one final project in which students may conduct researches of network science or conduct surveys of emerging techniques. Team collaboration is encouraged.

Paper(s):

report of the final project. oral presentation of the final project results required for on-site students.

Midterm Exam:

None

Final Exam:

None

Grading:

Homework assignments 50%, Final Project 50%

Hardware
requirements:

PC with Internet access. 

Software
requirements:

Students may use their preferred software (C, C++, Java, Matlab, R, SPSS) on their computers to complete homework assignments. 

Homework
submission:

by email to TA

Course Outline

Class Date

Class 
Number

Topics Covered

Assignment

Due

09/09/13

1

Overview of Network Science

 

 

09/16/13

2

Network Representations and Characteristics

HW #1

09/23/13

3

Network Partitioning and Visualization

09/30/13

4

Network Analysis Use Case

HW #2

HW #1

10/07/13

5

Network Sampling, Estimation, and Model

10/14/13

6

Network Topology Inference

HW #3

HW #2

10/21/13

7

Network Info Flow

10/28/13

8

Dynamic & Probabilistic Network and Graph Database (I)

HW #3

11/04/13

NO CLASS -- University Holiday

11/11/13

9

Final Project Proposals

11/18/13

10

Graph Database II

 

 

11/25/13

11

Knowledge Graphs

 

 

12/02/13

12

Impact of Network Analysis

 

 

12/09/13

13

Large-Scale Network Analysis System

 

 

12/16/13

14

Final Project Presentation