ICSI660 Anomalous Pattern Detection

Instructor

Feng Chen

Office

LI-96J

Number

(518) 442-4270

Email

fchen5@albany.edu

Office Hour

Wednesday: 11AM to 12PM

Thursday: 11:00AM to 12:00PM


Class Time and Location

Wednesday 5:45PM to 8:45PM

Class Website

http://www.cs.albany.edu/~fchen/course/2018-ICSI-660/


TA

Karan V Upadhyay

Office


Number

 

Email

kupadhyay@albany.edu


Course Description: 

This research seminar course introduces state-of-the art algorithms on the detection of anomalous or emerging events and other relevant patterns in the mobile context and/or data mining of spatial temporal, textual, or social media data. Examples of applications include disease outbreaks detection using public health data, such as hospital visits and medication sales; detection and prediction of crime events using historical crime record and streaming twitter data; and crowdsourcing human mobility and social media data to detect traffic congestion, air pollution, and power leakage events. More applications include the detection of computer network viruses, computer intrusions, malicious android applications, faked receipts and financial documents, Twitter robot accounts, new business hotspots, human right violations, Healthcare fraud activities, and etc. 

The first 10 class sessions will be lectures, and the following class sessions will be mainly student presentations and discussions. There will be no exams. The grade distribution is as follows: 

  1. Presentation: 20% + Class Discussion and Participation: 10%
  2. Option 1: Homework Assignments: 40% + Final Project (at most 2-member team): 30% 
  3. Option 2: Homework Assignments: 70% 

Homework Requirement: 

  1. Option 1: Correctly answer 50% of questions 
  2. Option 2: Correctly answer 90% of questions 
  3. You have two chances to make corrections of your homework answers. 


Readings

http://www.cs.albany.edu/~fchen/course/2018-ICSI-660/readings.html

Project Requirement:

There are three options: 

  1. Write a survey paper on anomalous pattern detection techniques in a specific domain (e.g., social media, computer network, transportation, finance).
  2. Implement one application of anomalous pattern detection on a real world data set (e.g., event detection in Twitter data, emerging resturants detection in Twitter or FourSquare data). 
  3. Design new anomalous pattern detection techniques and validate the proposed new techniques in simulation or real world datasets. 

Each team is allowed to have up to 2 members. The final project report must have at least 10 pages for teams of size 2 (6 pages for teams of size 1) and follow IEEE two-column style format, with the font name Times New Roman and size 10.

Submission deadline: To be announced. 

Schedule of Lectures

Week

Date

Lecture Topics

Reading Materials

1

8/29

Lecture 1: Introduction


2

9/5

Lecture 2: Passion distribution; normal distribution


3

9/12

Lecture 3: Maximum Likelihood Estimation


4

9/19 

(classes suspended)


5

9/26

Lecture 4: Hypothesis Testing


6

10/3

Lecture 4: Hypothesis Testing


7

10/10

Lecture 5: Anomalous Pattern Detection


8

10/17

Lecture 5: Anomalous Pattern Detection


9

10/24

Lecture 5: Case Studies


10

10/31

Lecture 6: Optimization


11

11/7

Lecture 6: Optimization


12

11/14

Lecture 6: Case Studies


13

11/21 

(classes suspended)


14

11/28

Student Presentation


15

12/5

Student Presentation


16

12/12 

Course Project Presentation



Student Presentations 

Student Name

Paper Title


Late Submission Policy: 

Assignments must be submitted before the class on the specified due date (Monday of designated week). A penalty of 30% will be deducted from your score for the first 24-hour period if your assignment is late. A penalty of 70% will be deducted from your score for >= 24-hour period. Assignments submitted more than 3 days late will not be assessed and will score as a zero (0). Weekend days will be counted. For assignments, you are encouraged to type your answers. 

Policy on Cheating: 

Cheating in an exam will result in an E grade for the course. Further, the students involved will be referred to the Dean's office for disciplinary action.

Homework problems are meant to be individual exercises; you must do these by yourself. Any of the following actions will be considered as cheating.

  1. A solution which is identical to or nearly identical to the solution submitted by another student in the class.
  2. A solution which is identical to or nearly identical to the solution provided by the instructor in a previous offering of CSI 660
  3. A solution which is identical to or nearly identical to a solution available on the Internet.

Cheating in a homework exercise will result in the following penalty for all the students involved.

  1. The homework in which cheating occurred will be assigned a grade of ZERO.
  2. The homework in which cheating occurred will be assigned a grade of ZERO.

Students who cheat in two or more homeworks will receive an E grade for the course. The names of such students will also be forwarded to the Dean's office for disciplinary action.

Attendance:

Class attendance is required and checked. Each case of missing class without a proper explanation will cause 20% less from your final numerical grade. If you miss a class, it is your responsibility to find out the material covered in the class. There will absolutely no makeup classes. Only in specific, unavoidable situations students are allowed to excuse absences from class: 1) personal emergencies, including, but not limited to, illness of the student or of a dependent of the student, or death in the family [Require doctor's note]; 2) religious observances that prevent the student from attending class; 3) participation in University-sponsored activities, approved by the appropriate University authority, such as intercollegiate athletic competitions, activities approved by academic units, including artistic performances, academic field trips, and special events connected with coursework; 4) government-required activities, such as military assignments, jury duty, or court appearances; and 5) any other absence that the professor approves.